Short-term Impacts of Tillage and Fertilizer Treatments on Soil and Root Borne Nematodes and Maize Yield in a Fine Textured Cambisol

An experiment was conducted at Chinhoyi University of Technology (CUT) farm, Zimbabwe. The experimental plots were established in December 2012, and the present study commenced in December 2014. The experimental station is located in a sub-tropical environment (17^○20′ S, 30^○14′ E, 1140 m above sea level) about 5 km east of Chinhoyi town. Soils are fine-textured Cambisols with a pH of about 5.6, mean daily temperature is 20.6°C with mean maximum temperature during summer of 27^○C and mean minimum temperature during the cold, dry season of 7^○C. Rainfall is unimodal, traditionally falling between November and April with a mean annual precipitation of 850 mm. The rainfall during the study period was 844.3 mm in 2014/15 and 679.7 mm in 2015/16 (Table 1).

A split plot randomized complete block design with three replications was used for this study. The main plot factor was tillage system: (i) rip line seeding (RIP), (ii) basin planting (BASIN), and (iii) conventional tillage (CONV), with fertilizer rate as the sub-plot factor (Table 2). RIP and BASIN represented conservation agriculture (CA) treatments whilst plots in which disc plowing was done represented conventional tillage treatments.

Crop residues were removed after harvesting in the CONV plots and retained at about 30% soil surface cover in CA plots. Basins were made using hand-held planting hoes; each basin measuring about 15 cm length × 15 cm width × 15 cm depth (Thierfelder and Wall, 2010). Some 5 cm of soil were added to the basin after placing the fertilizer, then another 5 cm of soil were added to cover the seed and the planting basins were about 5 cm deep after planting. Basin planting is a minimum tillage system that is commonly used by smallholder farmers who practice CA in Zimbabwe. Rip lines were made to a depth of about 10 to 15 cm using a tractor-mounted ripper. Conventional tillage was done using a disc plow as primary tillage followed by secondary tillage using a disc harrow. Basal fertilizer was applied at planting, about 5 cm below and 5 cm beside the crop seed. Nitrogen top dressing fertilizer (ammonium nitrate, 34.5% N) was applied 4 to 6 weeks after crop emergence depending on soil moisture availability. Most smallholder farmers in Zimbabwe apply 100 to 200 kg ha⁻¹ of basal fertilizer (7% N: 14% P₂O₅: 7% K₂O) and 80 kg ha⁻¹ of ammonium nitrate fertilizer (Kamanga et al., 2001). Both tillage and fertilizer treatments were applied on the same plots from the time the experiments were set up in 2012 and during each maize-growing season. Weeding was done four times during each cropping season using a hand-held hoe – no herbicides were used for weed control. Most farmers who practice CA in Zimbabwe use the hand hoe for controlling weeds (Muoni and Mhlanga, 2014).

Individual sub-plots measured 7.2 m (8 rows) × 8 m (32 plants per row) separated by 1.8 m pathways. Planting was done with the first effective rains in December of each maize-growing season. In all treatments, a medium maturity (130 d to maturity) commercial maize variety ZAP61 was used as the test crop, planted at four seeds per plant position and thinned to two plants per plant position three weeks after crop emergence at a plant spacing of 0.9 m inter-row by 0.5 m intra-row to give a target plant population of 44,444 plants ha⁻¹. Maize was grown on the same plots during the two cropping seasons of this study. Measurements were taken from the four central rows of 6 m in length.

Five soil sub-samples from a soil depth of 0 to 20 cm were randomly taken from each sub-plot using a 7.4-cm-diameter field core sampler to make one representative composite soil sample (250 cm³). Sampling was done at planting in December, then at 60 (February) and 120 d (April) after crop emergence (DAE). Three maize plants were also dug and uprooted from each sub-plot at 60 and 120 DAE and the stem cut off at the root crown level. The roots from these three maize plants were used to extract root borne nematodes whilst the above ground plant parts were discarded. Both soil and root samples from each sub-plot were bulked and thoroughly mixed, sealed in a plastic bag and stored in a refrigerator at 4°C until processed for nematode extraction. Sampling was done on the same set of plots during the crop growing season between November and April in 2014/15 and 2015/16 cropping seasons.

Nematodes were extracted from 100 cm³ of soil using the modified sugar centrifugation procedure and from 10 g of root tissue using the maceration centrifugation procedure (Southey, 1986) within two weeks of sampling. Plant-parasitic nematodes were identified to genus level and counted at 40 times magnification with a microscope. Non-plant-parasitic nematodes were all classified as “non-parasitic nematodes” without further assigning them to taxonomic groups. Classification was based on taxonomic characteristics, mainly from morphology and comparative anatomy (Bird and Bird, 1991). The main morphological characteristics used were: (i) body size and death posture, (ii) shape of head and cephalic region, (iii) stylet shape and size, (iv) esophagus and its overlap on the intestines, (v) position of vulva and number of ovaries in case of female nematodes and shape and position of bursa in case of male nematodes, and (vi) shape of tail. The plant-parasitic nematodes were further assigned to feeding site groups based on Yeates et al. (1993).

Maize grain was harvested from the sub-plot after physiological maturity and dried down in the field, in April, in each maize-growing season. Maize was harvested from 10 randomly selected samples of two rows, each 5 m long, collected from the middle of each plot and grain yield adjusted to 12.5% moisture content and expressed per hectare. All maize stalks and leaves without cobs were weighed at harvest; 10 sub-samples per plot were air dried at least four weeks before the final dry weights were taken and biomass was calculated on an area basis. The rest of the biomass was returned to the field as surface mulch.

Residuals were tested for normality and homogeneity of variances using Shapiro–Wilk’s test for normality and Bartlett’s test for homogeneity of variances, respectively, and were found to require transformation. Nematode count data were therefore log (x + 1.5) transformed to normalize the variances before being subjected to statistical analyses. Plant-parasitic nematode richness, diversity, and evenness (Shannon, 1948; Shannon and Weaver, 1949) were calculated using Paleontological Statistics (PAST) package version 3.14 (Hammer et al., 2001) and used as measures of community composition. Repeated measures in a split plot design were used to analyze the two-year nematode data using the residual maximum likelihood (REML), linear mixed model procedure using GenStat Release for Windows Version 10 (VSN International, 2011). The model included the two treatments and their respective interactions as fixed effects (Fixed model: Constant + Block + Sampling time + Tillage + Fertilizer + Sampling time × Tillage + Sampling time × Fertilizer + Tillage × Fertilizer + Sampling time × Tillage × Fertilizer). The sampling time and all its interactions were considered as random effects, and the results presented in this study are the estimated average values for each year. Yield data for each year were subjected to the split plot analysis of variance using GenStat Release for Windows Version 10 (VSN International, 2011). Nematode counts were used as a covariate in order to better estimate treatment effect on maize grain yield. Treatment means were compared using Fisher’s least significant difference (LSD) test at 95% level of significance. Plant-parasitic nematode composition data for each season were subjected to principal component analysis (PCA) using CANOCO 5 (ter Braak and Smilauer, 2012). Nematode populations recovered from soil samples at planting and combined soil and root samples at 120 d after crop emergence were subjected to correlation analyses using PAST version 3.14 (Hammer et al., 2001).

About 79,775 nematode specimens were identified combined for all seasons, years, and treatments. These nematodes were composed of non-plant-parasitic nematodes (5,662 specimens) and eight genera of plant-parasitic nematodes (74,113 specimens). Plant-parasitic nematodes recovered in the present study belonged to five groups according to feeding site: sedentary endoparasitic (Meloidogyne and Rotylenchulus), migratory endoparasitic (Pratylenchus), semi-endoparasitic (Scutellonema and Helicotylenchus), ectoparasitic (Trichodorus and Xiphinema) and algal, lichen and moss feeding (Tylenchus). The root lesion nematodes (Pratylenchus spp.) were most predominant (70.5%) followed by Scutelonema (14.6%), Helicotylenchus (7.3%), and non-plant-parasitic nematodes (7.1%) whilst combined abundances of Rotylenchulus, Tylenchus, Meloidogyne, Trichodorus, and Xiphinema were only 0.4% of total nematode abundance (Fig. 1).

Figure 1.

There were no treatment-by-sampling time point interactions. Mean densities of nematodes that were extracted from the soil were significantly affected by tillage while those recovered from maize roots did not respond to tillage treatments (Table 3) in both cropping seasons. In 2014/15 cropping season, population densities of Helicotylenchus were higher under rip line seeding and basin planting where crop residues were left on the soil surface compared to conventional tillage where maize residues were removed before plowing (Table 3). In the same cropping season, total nematode (plant-parasitic and non-parasitic combined) population densities in the soil were higher under basin planting compared to rip line seeding and conventional tillage. However, tillage had no effect on Helicotylenchus and total nematode population densities in 2015/16 cropping season. Also, in 2015/16 cropping season non-plant-parasitic nematodes were the lowest under rip line seeding but were not affected by tillage in 2014/15 season (Table 3). Population densities of Scutellonema in the soil were consistently higher in basin planting and rip line seeding compared to conventional tillage in both 2014/15 and 2015/16 cropping seasons (Table 3). Based on specimens recovered from both soil and root samples, the effects of tillage on the populations densities of Pratylenchus, Rotylenchulus, and Tylenchus were not significant (Table 3).

When nematodes were classified into feeding site groups, semi-endoparasites were almost double in rip line seeding and triple in basin planting compared to conventional tillage in both 2014/15 and 2015/16 cropping seasons (Table 4). However, tillage had no effect on algal feeders, ectoprasites, sedentary endoparasites and migratory endoparasites. In 2014/15 cropping season, plant-parasitic nematode richness was significantly lower under rip line seeding compared to basin planting, but was not affected by tillage in 2015/16 season (Table 5). In contrast, tillage had no effect on plant-parasitic nematode evenness and Shannon indices.

Fertilizer had no effects on densities of nematodes at both generic and feeding site group levels. Furthermore, fertilizer had no effect on all the three diversity parameters (richness, evenness, and Shannon–Weaver indices) during the two cropping seasons of this study. There was no tillage × fertilizer interaction in both 2014/15 and 2015/16 cropping seasons.

In general, maize grain yield increased with decreasing intensity of mechanical soil disturbance in both seasons (Table 6). Maize grain yield was higher under BASIN (85%) and RIP (74%) than CONV during 2014/15 cropping season. In 2015/16 cropping season, maize grain yield was higher in BASIN (111%) and RIP (63%) than CONV. Fertilizer application had no effect on maize grain yield in both 2014/15 and 2015/16 cropping seasons (Table 6).

Projection of the principal component analysis (PCA) showed some association among most nematode genera except Pratylenchus (Figs. 2, 3). Axis 1 explained 54.33% and 72.5% variability in 2014/15 and 2015/16 cropping seasons, respectively. On the other hand, Axis 2 explained 28.18% and 16.17% variability in 2014/15 and 2015/16 cropping seasons, respectively. In both PCA diagrams, the first axis separates conventional tillage (CONV) plots from minimum tillage (BASIN and RIP) plots. The second axis distinguishes unfertilized (NF) plots from fertilized (LF, MF, and HF) plots. The association between fertilizer application rate and nematode genera was not consistent during the two cropping seasons. In both seasons, most nematodes, except Pratylenchus, showed an affinity for the basin planting treatment, with semi-endoparasites (Helicotylenchus and Scutellonema) and non-plant-parasitic nematodes being more closely associated with the treatment. Correlation analysis showed that during 2014/15 cropping season, maize grain yield was significantly and positively correlated with pre-planting populations of Helicotylenchus and total plant-parasitic nematodes (r = 0.51 and r = 0.37, respectively; Table 7). However, these correlations were not evident 120 DAE (2014/15) and throughout 2015/16 cropping season. Maize grain yield was not significantly correlated with Pratylenchus and Scutellonema in both 2014/15 and 2015/16 cropping seasons (Table 7).

Figure 2.

Figure 3