Comparison of medium-term organic and inorganic fertiliser application on leaf nutrient concentration and yield of maize in rural agriculture in the Mbizana area, Eastern Cape province, South Africa

Abstract

A five-year study was conducted in the Mbizana area, Eastern Cape province, to assess the seasonal and medium term effects of chicken manure and inorganic fertiliser application to an acid clay loam topsoil on nutrient element uptake and yield of maize in a field trial under dryland. Treatments consisted of a once-off application of three levels of dolomitic lime, annual application of two levels of inorganic fertiliser, and three levels of chicken manure. A no-till practice with hand planting and fertiliser application was used. Leaf and soil sampling, as well as harvesting were performed and analyses done. Chicken manure application induced similar maize leaf macro- and micro-nutrient concentrations to the recommended inorganic fertiliser (F_rec) level and, in general, significantly higher values than the traditional fertiliser level (F_trad). Augmenting chicken manure with inorganic nitrogen led to significant increases in leaf nitrogen and phosphorus levels. Grain yields under chicken manure were similar to those under F_rec, but far superior to those under F_trad. It is concluded that under acid soil conditions the improvement of soil fertility, either by organic or inorganic fertiliser application, can result in maize grain yields of about 3 Mg ha^-1 under the prevailing rainfall conditions.

Keywords:

• chicken manure
• nutrient uptake
• soil fertility
• yield

Introduction

Soils in the Mbizana area of the Eastern Cape province, South Africa, developed from pre-Jurassic sedimentary rock or Jurassic dolerite under an annual precipitation of 900 to >1 000 mm. Similar to many soils in the tropics and subtropics, most of the soils are dominated by kaolinite, gibbsite, quartz and hydroxyl-interlayered vermiculite, i.e. minerals that contain very little or none of the major plant nutrients in their structures. The dominance of the low-activity kaolinite and gibbsite suggests that the soils could have undergone an intensive weathering spell in the past. The presence of quartz is hypothesised to be of aeolian origin, possibly reflecting desert loess, as well as due to its resistance to weathering (Bühmann et al. 2006a). The geochemical composition of the parent materials, natural acidification and leaching of nutrients under the prevalent relatively high rainfall, as well as exploitive farming activities, have given rise to unproductive soils in home gardens and fields, and even in large commercial croplands. Soil acidity and very low nutrient status, especially phosphorus (P), potassium (K) and calcium (Ca), are major productionlimiting factors in home gardens and fields of rural farmers (Mandiringana et al. 2005, Bühmann et al. 2006b).

A four-year study at the Indian Agricultural Research Institute, New Delhi, India, by Singh et al. (1986) indicates that nitrogen (N) and P concentrations in grain and straw of wheat, as well as grain yield, increased with a 10 Mg ha^-1 farmyard manure (FM) application. However, a combination of FM+fertiliser N (at levels of 40 and 80 kg ha^-1) led to larger increases of elemental contents and grain yield than when either were applied singly. From a study at the Punjab Agricultural University, Ludhiana, India, Bijay-Singh et al. (1997) report that by the third year of application poultry manure produced significantly more rice grain yield than the same rates of inorganic N. The authors found that poultry manure augmented with inorganic N produced yields intermediate between the yields from the two sources applied alone. In a study by Mkhabela and Materechera (2003) in the Midlands of KwaZulu-Natal province, South Africa, the researchers found that the majority (69%) of farmers indicated improved soil conditions, as well as growth and yield of crops, after applying manure. The respondent farmers also indicated that, although yields were higher with inorganic fertiliser compared to manure alone, the combined effect of manure and fertiliser produced even higher yields. Compared to inorganic N application (60 and 120 kg ha^-1), in a field experiment at the Domboshawa Training Centre, Harare, Zimbabwe, Nyamangara et al. (2005) reported poor N recovery from cattle manure in the first season. However, total N uptake by maize increased with increasing manure application. Increasing cattle manure application levels to 37.5 Mg ha^-1 led to an increase in maize grain yield of 800% over the control. These authors reported that the best results in terms of increased N uptake and grain yield were obtained with combined manure and inorganic N applications. Mkhabela (2006) lists several studies conducted in the Midlands of KwaZulu-Natal province in which poultry manure application, ranging between 5 and 22 Mg ha^-1, improved nutrient uptake, as well as growth and biomass yield of maize. In their study conducted near Alice, Eastern Cape province, South Africa, with sheep kraal manure, Mhlontlo et al. (2007) found that uptake of N and P in the leaves of a local Amaranthus accession increased with increased manure application rates. They reported that at low manure rates (<2.5 Mg ha^-1), plant heights and fresh matter yields were comparable to those in the control. Their findings suggest that at applications >2.5 Mg ha^-1, nutrient uptake, growth and yield were comparable to 150 kg NPK fertiliser ha-1 application. Comparing the effects of chicken manure (47 Mg ha^-1, N:P:K = 1.64:0.69:1.78%), charcoal and inorganic fertiliser (N:P:K = 85:75:100 kg ha^-1) on a highly weathered soil under grain sorghum and rice cropping, in a study conducted near Manaus, Amazonas, Brazil, Steiner et al. (2007) found that chicken manure significantly improved K and P nutrition in comparison to plants that received compost and/or inorganic fertiliser. Cumulative grain yield obtained with chicken manure application was 22 kg ha^-1 per kilogram of N, P or K applied, compared to 20 kg ha^-1 per kilogram of fertiliser N, P or K applied. Ruiz Diaz and Sawyer (2008) report from a field study conducted at the Iowa State University Research Farm, Boone, Iowa, USA, that increasing poultry manure application to supply 100 to 200 kg N ha^-1 significantly increased maize grain yield. Similar maize yields were obtained when comparing similar low (100 kg ha^-1) and high (168-200 kg ha^-1) N applications of poultry manure and inorganic fertiliser.

The objectives of the present study were to assess the seasonal and medium-term effects of organic and inorganic fertiliser application to an acid silty clay loam topsoil on element uptake and grain yield of maize. In a previous study (Beukes et al. 2012), the effects of liming and fertilisation on element uptake and grain yield of maize were discussed.

Materials and methods

Site description

A medium-term (1999/2000 to 2003/04) field trial with maize was conducted at Magusheni (30°51′ S, 29°37′ E, 1 098 m altitude) in the Mbizana local municipality of the Eastern Cape province, South Africa. Long-term mean annual precipitation is 836 mm, with 662 mm occurring between October and April (the active growing season). Long-term mean annual average temperature is 18.1 °C with a maximum of 26.1 °C during February and a minimum of 8.9 °C in July. The experimental soil was classified (based on a field survey and soil analyses) as a medium deep (90 cm) Magwa form (Connemara family) (Soil Classification Working Group 1991). The experimental site was never cultivated before 1999. Pre-trial soil sampling of the 0–30 cm and 30–60 cm depth increments was performed, using a grid of nine sampling points to cover a trial area of approximately 35 m × 45 m. Analysis values of selected soil chemical and physical properties are presented in Table 1.

Table 1:  Pre-trial chemical and physical properties of the experimental soil

Depth (cm)	pH		CEC (cmolc kg^-1)	Extractable acidity^a (cmolc kg^-1)	Acid saturation (%)	Organic C ^b (%)	Total N (%)	P^c (mg kg^-1)	K (cmolc kg^-1)	Ca (cmolc kg^-1)	Mg (cmolc kg^-1)	Na (cmolc kg^-1)	Sand (%)	Silt (%)	Clay (%)
	H2O	KCI
0-30	5.31	4.38	15.75	1.70	45.4	3.85	0.13	4.4	0.14	0.59	1.21	0.10	9.6	52.8	37.6
30-60	5.51	4.56	8.48	0.60	49. 7	1.13	0.05	4.4	0.05	0.18	0.30	0.10	11.3	36.0	52.7
^a 1 M KCl extractable acidity (H+Al)
^b Walkley-Black method
^c Bray 1 extraction

Treatments and measurements

A randomised complete block design with three replicates was used to apply the treatments comprising (1) three levels of dolomitic lime application (0, 5 and 10 Mg ha^-1) in 1999; (2) annual applications of two levels of inorganic fertiliser [traditional (F_trad): N:P:K = 16:24:16 kg ha^-1 (2:3:2 (22) mixture); commercially recommended (F_rec): 65:60:123 kg ha^-1 (2:3:2 (22) mixture + 90 kg limestone ammonium nitrate (LAN) ha^-1 + 166 kg KCl ha^-1)]; and (3) three levels of chicken manure (CM) [control, N:P:K = 64:61:101 kg ha^-1 (5 Mg ha^-1); CM+N: = 89:61:101 kg ha^-1 (5 Mg ha^-1 + 90 kg LAN ha^-1)]. Mean analyses (dry basis) for the CM were 1.71% N, 1.63% P and 2.69% K, respectively. The applied CM had a water content of 25%. The additional 25 kg N ha^-1 (90 kg LAN ha^-1) for the F_rec and CM+N treatments was applied as a top-dressing at six weeks after planting (i.e. when the plants were about knee height). In this study, the organic and inorganic fertiliser treatments are compared at the control lime level. The inorganic NPK fertiliser was band-placed in a furrow about 10 cm deep and covered with soil to a depth of approximately 5 cm before planting the seed and covering completely. The CM was broadcast at planting and manually worked in with a fork. Five rows of maize cv. Silver King (a locally popular white, open-pollinated cultivar) was planted in a 3.6 m × 10 m plot at a row spacing of 0.9 m and inter-row spacing of 30 cm for a targeted plant population of 37 000 plants ha^-1.

Sampling of the leaf below and opposite the ear in the three middle rows of each plot, excluding the end plants, was performed at about 100 days after planting (DAP; at flowering). The cobs of the three middle rows were harvested at about 180 DAP by hand, counted and weighed. Ten representative cobs per plot were threshed and the grain weighed. Grain moisture content was determined with a FARMEX Moisture Meter™. Final grain yield masses were expressed at 12.5% moisture content. Composite grain samples per plot were taken for chemical analysis.

Immediately after harvesting, soil samples of the 0-30 cm soil layer were taken with a standard soil auger as follows: at a randomly chosen point on the middle row, three soil cores were augered in a tight triangular fashion in such a way that the plant line, and hence the fertiliser line, would be sampled directly by one of the augerings. A further two 0–30 cm samples were taken at 45 cm distance on either side of the maize row, mixed with the on-the-row samples and a composite sample taken.

Weather data was obtained from the nearest automatic weather station operated by the ARC–ISCW, located at Intsingizi (30°54′ S, 29°54′ E; 974 m altitude), south-east of the town of Bizana. Manual rainfall recordings were also made at the trial site.

Weed and pest control

Prior to planting, all plots were hand hoed. Immediately after planting all planted plots were sprayed using a Knapsack sprayer with a mixture of Mamba 360 SL and Alachlor. Prior to top-dressing a preventative hoeing was done on all plots. After planting, all plots were treated with Kombat cutworm bait. The maize was treated twice for stalkborer: at about 15 cm plant height with Cypermethrin and again at about knee height using Kombat stalkborer granules. All applications were done in accordance with manufacturer instructions.

Laboratory analysis

Leaf and grain samples were oven-dried at 65 °C and milled to pass through a 0.05 cm sieve, whereas soil samples were air-dried and passed through a 0.2 cm sieve. Plant samples (leaves and grain) were analysed for selected elements in accordance with the procedures of Zasoski and Burau (1977) (digestion of P, K, Ca, Mg, Cu, Fe, Mn and Zn), Matejovic (1995) (total N) and Parviz et al. (1996) (determination of P, K, Ca, Mg, Cu, Fe, Mn and Zn). Soil samples were analysed for soil pH (H₂O; KCl), extractable acidity (H+Al), organic C (Walkley-Black), total N, P (Bray 1), K, Ca and Mg in accordance with standard laboratory procedures (The Non-Affiliated Soil Analysis Work Committee 1990, Parviz et al. 1996). Soil acid saturation was calculated by expressing extractable acidity as a percentage of all measured cations.

Statistical analysis

The data for the five seasons were pooled per variable and an analysis of variance performed using the GenStat statistical software package (GenStat 5 Committee 1993). Fisher variance ratios were calculated and expressed as probabilities (p). Fisher's least significant difference (LSD) procedure (Snedecor and Cochran 1967), with a level of significance of p = 0.01, was used to identify differences between treatment means.

Results and discussion

Soil characteristics and rainfall during the study period

The A horizon (0–30 cm) exhibited pertinent acid conditions with low soil pH and basic cation values, as well high extractable acidity (H+Al) and acid saturation (AS) values (1.7 _cmolc kg^-1 and 45%, respectively) (Table 1). The low P and basic cation values are probably indicative of typical ‘acid soil infertility’. Notable is the high organic C value (3.9%) in the top 30 cm, which is probably the reason for the relatively high (FSSA 1988) cation exchange capacity (CEC) value, and also leads to a humic horizon classification. The B horizon is also very acidic, especially in terms of an AS of 50%, which is primarily caused by extremely low basic cation values. Texturally, the A and B horizons can be classified as silty clay loam and clay, respectively (Table 1). Seasonal (October–April) rainfall was very variable: 410 mm (1999/2000), 568 mm (2000/01), 820 mm (2001/02), 482 mm (2002/03) and 618 mm (2003/04). Mean seasonal rainfall was 580 mm, compared to the long-term mean of 662 mm. Monthly rainfall peaked in January (110 mm), with a typical midsummer drought during February with a mean of only 50 mm (data not included).

Nutrient concentration and grain yield

Table 2 shows that seasons and treatments significantly affected the maize leaf macro- and micro-nutrient concentrations, viz. N, P, K, Ca, Mg, Cu, Fe (seasons only), Mn and Zn, as well as grain yield. Except for Fe, significant season × treatment effects were also noted for the former parameters. The recommended inorganic fertiliser application (F_rec) induced significantly higher leaf N uptake (23.96 g kg^-1) compared to CM and F_trad. A notable result is that chicken manure augmented with inorganic N (CM+N) gave a statistically similar N uptake to F_rec. Similar results have been reported elsewhere (e.g. Singh et al. 1986, Mkhabela and Materechera 2003, Nyamangara et al. 2005). The lower N recovery rate by the maize with CM, compared to F_rec, at similar application rates (64–65 kg N ha^-1) could be ascribed to the slower mineralisation rate of organic N in the manure (Table 2, Figure 1a). The dynamics of manure utilisation, decomposition, as well as the contribution of manure to soil fertility, have been extensively discussed by Mkhabela (2006). The season × treatment interactions depicted in Figure 1a show large variations that could be caused by the variable seasonal rainfall. Leaf N uptake for both organic and inorganic fertiliser application was much higher than the control leaf N of 15.5 g kg^-1 (Figure 1a). The latter value, as well as leaf N concentrations in some of the seasons, were below the mean deficiency value of 22.9 g N kg^-1 of Reuter and Robinson (1997).

Figure 1:  Fertiliser effects on leaf nutrient concentration and maize grain yield in each growing season. (a) Nitrogen, (b) phosphorus, (c) potassium, (d) calcium, (e) zinc and (f) grain yield

Table 2:  Statistical parameters for plant properties

Source of variation	F ratios and probabilities (p)
	Maize leaf elements
	N	P	K	Ca	Mg	Cu	Fe	Mn	Zn	Grain yield
Season (S)	28.5^***	12.1^**	6.1^*	24.0^***	13.6^***	24.1^***	54.4^***	17.8^***	22.4^***	22.5^***
Treatment (T)	97.7^***	84.0^***	105.6^***	12.2^***	46.3^***	28.2^***	0.5^ns	20.0^***	13.3^***	68.8^***
S×T	22.0^***	6.0^***	4.3^**	2.6^*	3.7^*	6.2^***	0.7^ns	5.1^***	3.0^**	2.5^*
Treatment	Treatment means
	N(g kg^-1)	P(g kg^-1)	K(g kg^-1)	Ca(g kg^-1)	Mg(g kg^-1)	Cu(g kg^-1)	Fe(g kg^-1)	Mn(g kg^-1)	Zn(g kg^-1)	Grain yield(g kg^-1)
CM	21.51^b	2.19^ab	14.84^b	3.75^a	4.85^b	8.5^a	188^a	44.3^b	21.2^ab	2.905^a
CM+N	23.11^a	2.31^a	15.48^b	3.62^a	4.37^b	8.9^a	208^a	48.1^ab	23.5^a	3.104^a
Ftrad	17.43^c	1.39^c	10.93^c	3.90^a	6.0^a	6.3^b	225^a	34.9^c	18.8^b	1.303^b
Frec	23.96^a	2.13^b	16.60^a	3.21^b	3.92^c	9.0^a	208^a	52.2^a	18.5^b	2.880^a
LSD (p = 0.01)	1.15	0.18	0.94	0.33	0.52	0.9	86	6.5	2.5	0.407
Respective treatment column means followed by different letters are significantly different ^* p = 0.05, ^** p = 0.01, ^*** p = 0.001, ns = not significant

Application of just chicken manure (CM) gave a similar leaf P concentration to F_rec (2.19 vs 2.13 g kg^-1), whereas CM augmented with inorganic N gave significantly higher leaf P than both inorganic fertiliser applications (Table 2, Figure 1b). Similar results were reported by Singh et al. (1986), Steiner et al. (2007) and Mhlontlo et al. (2007). The mean seasonal leaf P value of 1.4 g kg^-1 for the traditional fertiliser application (F_trad) level (Table 2) is similar to the control value of 1.43 g kg^-1 (Figure 1b). Both values are below the deficiency level of 1.6 g P kg^-1 quoted by Reuter and Robinson (1997). The highest K application level (F_rec: 123 kg ha^-1) resulted in significantly higher K uptake compared to all the other treatments. Notable is the significantly higher K uptake with the CM applications compared to F_trad (Table 2). Steiner et al. (2007) also reported improved K nutrition of grain sorghum and rice in response to chicken manure application. Except for F_trad, the season × treatment display of leaf K in Figure 1c shows that leaf K mostly fluctuated in the marginal concentration range of 12–16 g K kg^-1 of Reuter and Robinson (1997). F_trad fluctuated notably lower at 9–14 g K kg^-1.

As can be expected on an acid soil with a low Ca status (Table 1), leaf Ca concentrations (Table 2) were in the critical range of 2–4 g Ca kg^-1 as defined by Reuter and Robinson (1997). Leaf Mg concentrations were above the critical value of 2.5 g Mg kg^-1 (Reuter and Robinson, 1997), which could be ascribed to the fact that residual soil Mg was relatively high (Table 1) because of the fact that one of the parent materials, viz. Jurassic dolerite, is high in Mg (Bühmann et al. 2006a). The season × treatment display of leaf Ca in Figure 1d shows that, at least in the first two seasons, leaf Ca for all treatments was lower than the control value of 3.43 g Ca kg^-1. Organic and inorganic fertiliser application did not exhibit clear effects on micronutrient uptake (Table 2) and it would appear as if chicken manure application can contribute to leaf values similar to inorganic fertiliser. Notable is the significantly higher leaf Zn with CM + inorganic N compared to inorganic fertiliser. Figure 1e displays large seasonal variation in leaf Zn (not determined in 2001/02), a phenomenon that cannot be explained yet. The figure shows that for some fertiliser treatments and seasons, leaf Zn was below the control value of 17 g kg^-1, and sometimes below the critical value of about 15 g Zn kg^-1 of Reuter and Robinson (1997). Although not significant, CM application produced higher maize grain yields than F_rec, whereas CM augmented with inorganic N gave the best yield of 3.104 Mg ha^-1. Notable is the fact that CM performed significantly better than F_trad (Table 2: 2.905/3.104 vs 1.303 Mg ha^-1). These results are in agreement with other reports (e.g. Singh et al. 1986, Bijay-Singh et al. 1997, Mkhabela and Materechera 2003, Nyamangara et al. 2005). Figure 1f shows large seasonal variation in yield for all treatments, probably caused by variation in seasonal rainfall. For example, seasonal grain yields for CM+N varied between 2.438 and 3.850 Mg ha^-1, and those for F_trad between 0.199 and 1.847 Mg ha^-1. Comparing the grain yields obtained under organic and inorganic fertilisation with the control yield of 0.119 Mg grain ha^-1, it can be concluded that under acid soil conditions the improvement of soil fertility can result in maize grain yields of about 3 Mg ha^-1 under the prevailing rainfall conditions.

Conclusions and recommendations

Over a period of five growing seasons and under a mean seasonal rainfall of 580 mm, the application of chicken manure, chicken manure augmented with inorganic N, as well as inorganic fertiliser on an acid silty clay soil, significantly affected leaf macro- and micro-nutrient concentrations and maize grain yield. The season × treatment interactions probably indicate the influence of the observed variability of seasonal rainfall and other climatic factors. It does appear as if there is a lower N recovery rate by maize with chicken manure compared to inorganic fertiliser at similar application rates. This could be because of the application of CM at planting. An earlier application probably would have given a more comparable N recovery rate. Application of just CM gave a similar leaf P concentration to the inorganic F_rec, whereas CM augmented with inorganic N gave significantly higher leaf P than both inorganic fertiliser applications. Potassium uptake was significantly higher with the CM applications compared to the inorganic F_trad. As can be expected on an acid soil with a low Ca status (Table 1), leaf Ca concentrations were below, whereas leaf Mg concentrations were above, the critical values, which could be ascribed to the fact that residual soil Mg was relatively high. Organic and inorganic fertiliser application did not exhibit clear effects on micro-nutrient uptake, whereas the results indicate that CM application can contribute similar leaf values to inorganic fertiliser. Increasing liming and fertilisation levels significantly increased leaf macro-nutrient (N, P and K) concentrations. Leaf Ca and Mg were only increased by liming. Certain leaf micro-nutrient (Cu, Mn and Zn) concentrations either increased or decreased with liming and fertilisation application (Beukes et al. 2012). It can be concluded that the type and amount of lime and fertiliser applied will differentially affect leaf nutrient concentrations. It was found for the control maize plants that most leaf macro-nutrient concentrations were marginal to deficient, whereas micro-nutrient levels were in the adequate range. With CM and recommended inorganic fertiliser applications, leaf macro-nutrient concentrations were adequate to marginal, whereas micro-nutrient levels were adequate to high. With the traditional fertiliser application level, leaf macro-nutrient levels were mostly marginal to deficient. Chicken manure augmented with inorganic N gave similar or higher leaf nutrient concentrations compared to the recommended inorganic fertiliser level. Although not significant, CM application produced higher maize grain yields than the recommended inorganic fertiliser level, whereas CM augmented with inorganic N gave the highest yield. Notable is the fact that chicken manure produced significantly higher yields than the traditional fertiliser level. Beukes et al. (2012) found that, while liming had no clear effect on maize grain yield, increased fertilisation led to significantly higher grain yields. It can finally be concluded that under the acid soil conditions of the study area, the improvement of soil fertility, either by organic or inorganic fertiliser application, can result in maize grain yields of about 3 Mg ha^-1 under the prevailing rainfall conditions. The augmentation of manure with inorganic fertiliser to include all major nutrient elements should be investigated for the various available sources, and the results communicated to the local farmers.

Acknowledgements

The authors gratefully acknowledge the funding of the project by the Eastern Cape Department of Agriculture and Land Affairs (ECDALA) under the auspices of the National LandCare Programme of the Department of Agriculture, Forestry and Fisheries. Sincere thanks are extended towards Chief Ntola for permission to use the land at Magusheni, extension staff of ECDALA for their participation, Mr Angus Judge of ARC–ISCW and Mr Peter Wood of KeyPoint RDSS Trust, for their ever-willing assistance with, and dedication to, the field trial activities. Thanks also go to Mrs Marie Smith of the ARC–Biometry Unit for assistance with the experimental design and statistical analyses.