Introduction

Sweetpotato (Ipomoea batatas) is the world’s sixth most important food crop, with more than 105 million metric tonnes produced globally each year. Although sweetpotato originated in central and south America, Asia now accounts for 86% of the production, with China by far the largest producer. Australia is a relatively small contributor, with total production of around 100,000 t. Nevertheless, its sweetpotato industry has grown remarkably in recent years, with sales increasing by about 20% per annum and yearly production now valued at more than $80 million. Most of the crop is grown in subtropical regions, with Bundaberg (southern Queensland) and Cudgen (northern New South Wales) the main centres of production.

Worldwide, sweetpotato is damaged by a wide range of insect and nematode pests, but because storage roots are the marketable product, pests which damage these roots and render them unfit for sale are the most damaging from an economic perspective. Root-knot nematodes (Meloidogyne spp.) are perhaps the most important of these pests because they occur in almost all regions where sweetpotatoes are grown and severely affect both yield and quality (Clark et al. 2013). Effects on yield are due to a reduction in root elongation, which limits water and nutrient uptake. However, effects on marketability are even more pronounced. Root-knot nematodes invade roots and grow to maturity within storage roots, causing uneven protuberances on the root surface and necrotic lesions within root tissue. In some cultivars, longitudinal cracks are also produced on storage roots (Clark et al. 2013).

The genus Meloidogyne contains more than 100 species but in Australia only four species (M. javanica, M. incognita, M. arenaria and M. hapla) are associated with sweetpotato. The first three species (sometimes referred to as ‘warm-climate species’) are widely distributed in coastal NSW and southern Queensland, particularly on sugarcane and various vegetable crops (Blair et al. 1999; Hay and Stirling 2014) whereas M. hapla is largely confined to elevated areas where temperatures are lower. Sweetpotato growers often report losses in marketable yield due to these nematodes and the Australian industry considers root-knot nematode by far its most important pest.

Currently, sweetpotato growers have a range of control options but they all have their limitations. Cultivars such as Bellevue are resistant to M. incognita (La Bonte et al. 2015) but it is not clear whether they are resistant to M. javanica, the most widespread root-knot nematode species in areas where sweetpotatoes are grown in Australia. Forage sorghum varieties with resistance to root-knot nematode (Stirling et al. 1996) are excellent rotation crops but in situations where limited land is available, it may not be possible to grow them for long enough to reduce nematode populations to non-damaging levels. Also, volunteer sweetpotatoes usually grow under the rotation crop and they always carry over some nematodes.

Most sweetpotato growers currently rely on nematicides for nematode control, with two soil fumigants (metham sodium and 1, 3 dichloropropene) and two non-volatile chemicals (oxamyl and fluensulfone) currently approved for use by the Australian Pesticides and Veterinary Medicines Authority. However, all nematicides are expensive, many have off-site environmental impacts, some lose their efficacy due to enhanced biodegradation, and broad-spectrum fumigants such as metham sodium destroy beneficial organisms as well as the target pest (Thomason 1987; Stirling et al. 1992; Warton et al. 2001; Matthiessen et al. 2004: Sánchez-Bayo 2011; Blaesing 2014).

Although better nematode management practices are required for sweetpotato, one issue that growers must also consider is the degraded state of their soils. Most of the land used for sweetpotato production has been cropped to sugarcane, vegetables or other crops for more than 100 years and soils with compacted layers, low moisture holding capacities, poor rainfall infiltration rates, low organic carbon levels and limited microbial activity are relatively common (Bridge and Bell 1994: Stirling 2008). From a biological perspective, frequent tillage and the regular use of pesticides and fumigants will have modified the soil biology to such an extent that the mechanisms that normally regulate populations of nematodes and other soilborne pests and pathogens are no longer likely to be operating effectively.

Given the above, one way the Australian sweetpotato industry could improve the health of its soils, overcome its root-knot nematode problems and have a more sustainable future is to adopt an approach that was termed ‘Integrated Soil Biology Management’ by Stirling (2014). The basis of this concept is that a range of nematode management practices are integrated into a farming system to not only reduce pest nematode populations but also improve the physical, chemical and biological health of the soil. This means that crop rotation and cover cropping must become integral components of the farming system, bare fallows are eliminated, tillage is minimised, crop residues are retained on the soil surface, levels of soil organic matter are raised with cover crops and organic amendments, compaction and other soil physical constraints are removed, inputs of nutrients are optimised, and pesticides are used judiciously. Evidence that the above-mentioned practices should be part of a sustainable farming system can be found in reviews such as Lehman et al. (2015) and Norris and Congreves (2018), while case studies showing how some Australian growers have integrated many of these practices into their farming system can be found in Stirling et al. (2016).

Over the last 40 years, a considerable amount of research on field crops has shown that when practices such as strategic crop rotation, retention of plant-residues on the soil surface and minimal tillage are used collectively, they will improve productivity without having a negative impact on soil health (Wolf and Snyder 2003; Franzluebbers 2004; Pretty 2008; Hobbs et al. 2008). Consequently, grain and oilseed growers in many countries now integrate these practices into their farming systems. Conservation agriculture (i.e. no-till planting systems with surface retention of crop residue and rotation of crops) is now standard practice in Australia and the reasons why it has been adopted and the many benefits obtained are covered in detail by Cornish and Pratley (1987) and Pratley and Kirkegaard (2019).

Conservation agriculture is not used to the same extent in vegetable production, as most vegetable-growing soils are repeatedly tilled. The farming system used by most sweetpotato growers in Australia is a typical example. Fields are tilled several times after harvest to kill weeds and eliminate volunteers, cover crops are incorporated with a rotary hoe, and rippers, disc-harrows and rotary hoe bed-formers are used to prepare beds for the next crop.

Because vegetable crops are managed more intensively than field crops and some soil disturbance is necessary when root crops such as sweetpotato are harvested, developing production systems that improve rather than deplete soil health is challenging. Nevertheless, improvements in soil health were obtained when cover cropping, reduced tillage and soil amendments were integrated into vegetable farming systems (Stirling et al. 2012; Stirling 2013) and the results of a recent meta-analysis (Norris and Congreves 2018) confirmed that these practices are beneficial. Sixty studies on a wide range of vegetable crops were included in the analysis and although all three practices improved soil health, the results indicated that organic amendments had the greatest impact. Various amendments were used, including composts, manures, biochar, yard wastes and other organic materials, and they decreased bulk density and increased soil porosity, soil moisture, soil carbon levels, microbial biomass and biological activity.

Another reason for integrating organic amendments into sweetpotato farming systems is to provide some control of root-knot nematode, thereby reducing the need for chemical nematicides. Organic amendments were first used to control plant-parasitic nematodes in the 1920s and there is now a huge body of evidence to indicate that they should be a component of integrated nematode management programs (see reviews by Muller and Gooch 1982; Akhtar and Malik 2000; Oka 2010; Thoden et al. 2011; McSorley 2011; Renčo 2013). However, determining the best way of using organic amendments is not a simple process because adding organic matter to soil initiates a series of transformations that directly or indirectly affect the soil physical and chemical environment, the soil biological community, and ultimately, plant growth (Stirling 2014). Thus, an amendment must be chosen that not only stimulates natural enemies and enhances the soil’s suppressiveness to nematodes, but also improves soil health and maintains or improves crop yields.

Background to this study

As mentioned previously, the farming system currently used to produce sweetpotatoes in Australia involves a huge amount of tillage. Given that tillage increases the soil’s susceptibility to wind and water erosion, reduces soil carbon reserves and is detrimental to the soil biota, we devised an alternative farming system that was expected to improve rather than degrade the health of the soil. This system incorporates the key principles of conservation agriculture and its five components are illustrated in Fig. 1. First, beds are reformed a few weeks after harvest so that the soil biological community can begin to recover from the disturbance caused by the harvesting operation. Second, an organic amendment is incorporated into the soil during the bed formation process to increase soil carbon levels and enhance biological activity. Third, two cover crops are grown (e.g. forage sorghum and oats) so that biomass production can be maximised in both the warmer and cooler months of the year. Fourth, the residues produced by the cover crops are retained on the soil surface as mulch. Fifth, the next sweetpotato crop is established using a strip till process in which a channel is formed in the middle of previously-prepared beds and stem cuttings are planted into that channel. This system was tested on-farm to see whether it had the potential to replace the farming system currently used by most sweetpotato growers in Australia. The methodology and our observations are reported in the first part of this paper.

Fig. 1
figure 1

Components of an alternative sweetpotato farming system. a. Bed re-formed soon after harvest and an organic amendment incorporated into the soil during the process. b. Forage sorghum cover crop. c. Cover crop residues retained on the soil surface as mulch. d. Second cover crop. e. Sweetpotato planted into mulched, undisturbed bed

Full size image

A second component of this study aimed to determine whether organic amendments could be used to reduce losses from root-knot nematodes. Research undertaken in various countries has shown that organic amendments are beneficial to sweetpotato, as they increase yields, enhance its health-promoting properties and improve post-harvest quality (Antonious et al. 2011; Siose et al. 2018; Atuna et al. 2018). Organic amendments also provide useful levels of nematode control, with a field trial in Nigeria showing that various animal manures, but particularly poultry manure applied at 10 t/ha, markedly reduced the reproduction of M. incognita and also reduced the severity of nematode damage to storage roots (Osunlola and Fawole 2015).

A wide range of organic materials are used as soil amendments in Australia (Quilty and Cattle 2011), but we chose to work with three organic wastes that would enable sweetpotato growers to contribute to a ‘circular economy’ in which agriculture’s fodder-manure cycle is closed and wastes are recycled. Chicken litter, sawdust and compost were selected because they are readily available in northern NSW and Queensland and there is some evidence to indicate that they reduce nematode populations and improve plant growth.

Chicken litter, a by-product of chicken meat production, is a mixture of manure and a bedding substance (usually sawdust, rice hulls or straw). More than one million tonnes of chicken litter are produced in Australia every year and a comprehensive report by Weidemann (2015) presents information on its availability, composition, and nutrient characteristics. It also provides guidelines on its use as a fertiliser and soil conditioner. The main reason chicken litter is used as an amendment is that it contains nitrogen, phosphorus and potassium and a full complement of micronutrients. However, it also contains significant quantities of carbon, and this means that it can be used in high-input horticultural industries to increase organic matter levels and improve soil health. As greenhouse tests have shown that chicken litter is detrimental to root-knot nematodes and that the level of control is directly proportional to application rate (Mian and Rodriguez-Kabana 1982; Kaplan and Noe 1993; Riegel and Noe 2000), it was the first amendment chosen for study.

Sawdust, a waste product of the saw mill industry, was chosen for several reasons. First, large quantities are available in Australia, and although sawdust and other wood residues are often combusted for power generation, they can be used to produce biochar or applied to soil as an amendment (Goble and Peck 2013). Second, waste products such as sawdust should be recycled or re-used rather than finish up in landfill. Third, sawdust reduces damage caused by root-knot nematode (Stirling 1989; Vawdrey and Stirling 1997) and is now widely used in the Queensland ginger industry to protect rhizomes from nematode attack and provide a range of other benefits (e.g. ease of digging, cleaner rhizomes and better retention of soil moisture) (R. Abbas and M Smith, pers. comm.).

Compost was included in this study because it is commonly used as a soil conditioner and supplies made from household wastes, animal manures and other organic materials are readily available. Growers can also produce compost on-farm using plant residues and various organic materials. When used as a soil amendment, composts have positive effects on soil structure, soil fertility and soil biology, and will usually improve plant growth and yield. However, a review of studies on the use of compost for nematode control found that it often increased populations of plant-parasitic nematodes and did not enhance the soil’s suppressiveness to nematodes (Thoden et al. 2011). Importantly, however, composts markedly increased populations of free-living nematodes, suggesting that the effects of these nematodes (particularly the bacterivores) on nutrient availability may have been the reason crop yields consistently increased. As bacterial-feeding nematodes improve the growth and vitality of plants, thereby reducing their susceptibility to plant-parasitic nematodes, the authors concluded that composts and other organic amendments should always be a component of integrated management programs for nematode pests.

The other organic material considered for inclusion in this study was sugarcane trash (i.e. the tops and leaves left in the field after a sugarcane crop is mechanically harvested). Large quantities of this material are available in northern NSW and Queensland, as about 25 t/ha of dry sugarcane residue is left on the soil surface when crops yielding 125 t/ha are harvested (Mitchell and Larsen 2000). Previous studies have shown that the soil under a blanket of sugarcane mulch is suppressive to plant-parasitic nematodes (Stirling et al. 2011a) and that populations of root-knot and root-lesion nematode are markedly reduced when cane-growing soils are amended with sugarcane trash (Stirling et al. 2003, 2005). Sugarcane residues are relatively easy to retrieve from cane fields and are often sold as mulch for home gardens, but they were not used in this study for two reasons. First, the material is light and bulky, so it is difficult to apply large quantities to soil, and second, removing crop residues from cane fields is detrimental to the long-term sustainability of the sugar industry.

Results from field trials and greenhouse tests in many countries have shown that when organic amendments are used for nematode control, application rates of 10–100 t/ha are often required (Thoden et al. 2011; McSorley 2011). As application rates at the higher end of this scale are never likely to be economically feasible in a crop such as sweetpotato, we felt that the best way to use amendments would be to apply them in a concentrated band around the developing storage roots, as they are the marketable product that is subject to nematode damage. We hypothesised that if amendments were applied to a V-shaped furrow in the centre of the bed using a double-disc opener and sweetpotato cuttings were planted in that furrow (Fig. 2), the storage roots would develop in a zone where biological activity was high enough to suppress populations of root-knot nematode. Results from a field trial established to test that hypothesis are presented in the second part of this paper.

Fig. 2
figure 2

Diagram showing an organic amendment placed in a V-furrow at the top of the bed, and sweetpotato planted in that furrow

Full size image

Materials and methods

Assessment of an alternative sweetpotato farming system

A farming system that incorporated the key principles of conservation agriculture was assessed in a field at Cudgen, NSW where sweetpotatoes had been grown for more than 20 years. The chosen field had a relatively fertile volcanic soil (Red Ferrosol according to the Australian Soil classification) and its most recent sweetpotato crop (harvested in November 2017) had suffered severe damage from M. incognita. The field was cultivated several times in the following month and then on 28 December 2017, chicken litter and sawdust were broadcast over an area 15 m wide and 100 m long and then incorporated into the soil with a rotary hoe. The application rate was determined by weighing the amount of organic material collected on four randomly-spaced plastic sheets each 1 m2 and this showed that the chicken litter and sawdust were applied at 72.2 and 43.9 t dry weight/ha, respectively. Analyses of the C and N contents of the two amendments indicated that their C/N ratios were 11.5 and 467, respectively.

On 7 February 2018, ten beds spaced 1.5 m apart were formed in the amended area. As two of these beds were to be the focus of this study (they were to be left undisturbed when the next sweetpotato crop was planted), four plots each 50 m long were marked out in these beds. Seeds of forage sorghum (Sorghum bicolor x S. sudanense cv. Sweet Jumbo LPA) were then sown in all of the beds, with rows of the cover crop spaced about 15 cm apart. In early May, when the forage sorghum was about 2 m high, the cover crop was slashed and the aboveground biomass left on the soil surface as mulch. A month later, the ratooning forage sorghum was sprayed with glyphosate and on 25 June seed of a second cover crop (oats, Avena sativa cv. Saia) was broadcast over the beds at a rate of 120 kg/ha. Winter rainfall was sufficient to germinate the seed and produce a good crop, and when it was terminated with glyphosate, fluazifop-P and an emulsifier on 30 September, the aboveground residues were also retained on the soil surface as mulch. As oats cv. Saia was known to be susceptible to M. incognita (Stirling et al. 1996), sixteen randomly selected plants were collected on 20 November 2018, their root systems were placed in a mist cabinet for 7 days and the nematodes extracted were checked for the presence of root-knot nematode.

In mid-November, eight of the ten beds were prepared conventionally for planting sweetpotato (i.e. a rotary hoe was used to incorporate residues from the rotation crops and a bed-former was then used to reform the bed). The other two beds were left as they were so that the performance of sweetpotato in undisturbed, mulched, organically-amended beds could be assessed. Sweetpotato cv. Orleans was planted in all beds on 3 December 2018.

The sweetpotato crop was harvested on 5 July 2019 (31 weeks after planting) and yield was assessed by harvesting a randomly-selected 1 m section in each of the four plots. Between 25 and 40 storage roots were retrieved from the harvested plots and they were rated individually for root-knot nematode damage using the following scale: 0 = no symptoms; 1 = a few raised blisters or pimples on the surface of the root; 2 = pimples readily visible but only in localised areas of the root, and necrotic lesions, when present, relatively sparse and confined to surface tissue; 3 = large numbers of pimples covering most of the root surface, with lesions often extending several mm into the root tissue, and significant nematode damage at the distal end of the storage root; 4 = similar symptoms to rating 3, but longitudinal cracks also present. As roots with a damage rating of 3 or 4 were considered unmarketable, they were then separated from the others and the weights of marketable and unmarketable roots were obtained.

At various times during the 18-month study, a sampling tube 22 mm in diameter was used to collect 15 soil cores to a depth of 20 cm from random positions in each plot. Each sample was then mixed gently, a 200 mL sub-sample (about 230 g moist weight) was spread on a standard extraction tray (Whitehead and Hemming 1965) and nematodes were recovered after 2 days by sieving twice on a 38 μm sieve. Numbers of root-knot nematode and total numbers of free-living nematodes were counted at a magnification of 40X. In the samples that were collected when the sweetpotatoes were harvested, the composition of the free-living nematode community was also assessed. A random sample of 150–200 nematodes was examined under a microscope, the proportion in various trophic groups was determined, and the Nematode Channel Ratio (Yeates 2003) was calculated as B/(B + F), where B and F are the number of bacterivores and fungivores, respectively.

At three of the sampling times, some of the soil in each sample was air-dried and various chemical and biochemical parameters were assessed. The Dumas dry combustion method was used to measure total organic C and total N, while permanganate-oxidisable C was analysed using the procedure of Blair et al. (1995), except that the permanganate concentration was reduced to 33 mM. Soil microbial activity was assessed using the Solvita® CO2-burst protocol (Haney et al. 2001; Franzluebbers 2016). Beakers were filled with 40 g of air-dried soil, the soil was moistened with 20 mL of demineralised water and the flush of CO2 was measured after 24 h using a digital colour reader.

Field trial with organic amendments in a V-furrow

This experiment was set up in the same field as the alternative farming system study described above. The previous sweetpotato crop was due to be harvested in November 2017, but damage from M. incognita was so severe that the crop was abandoned and the field cultivated several times with a rotary hoe. Soil samples collected in early December 2017 indicated that the root-knot nematode population was about 950 nematodes/200 g dry weight soil. In December 2017 and January 2018, weeds and volunteer sweetpotatoes were killed by tilling the soil with a rotary hoe and also applying herbicides (glyphosate and MCPA). Then on 7 February 2018, ten beds were formed at the standard bed spacing of 1.5 m while the remaining area was handled in the conventional manner (i.e. it was not re-bedded). The site was then cover-cropped with forage sorghum and oats, as previously described. Beds for the next sweetpotato crop were formed in mid November 2018 and during that process, the conventionally managed area was deep-ripped, disced with disc harrows, rotary hoed twice and then a rotary hoe bed-former was used to prepare the beds. Cultivation was less aggressive in beds that were already in place because only two passes with the bed-former were required.

Four rows were used for the experiment: two rows where beds were formed in February and two rows where beds had been prepared conventionally in mid November. The nematode population was assessed on 19 November 2018 by collecting six soil samples (each 15 cores to a depth of 20 cm) from both the early and late-formed beds and extracting the nematodes from 460 g moist wt. soil using the method described previously. An experiment consisting of two bed formation times (early and late) and four organic amendment treatments (nil, sawdust, compost and sawdust/chicken litter) was then established. The intention was to apply each amendment at a rate of 40 m3/ha and because there are 6667 m row/ha at a bed spacing of 1.5 m, this meant that the amendments were applied at a rate of 6 L/m of row. This was achieved by digging a V-shaped furrow about 14 cm deep and 9 cm wide in the middle of the bed and filling it with the amendment. As each plot was 8.3 m long, 50 L of amendment was applied per plot. The amendments were compost made from yard waste, sawdust obtained from a pine sawmill and chicken litter from a meat chicken production facility. The sawdust-chicken litter amendment was a mixture containing 60% sawdust and 40% chicken litter (v/v). On a dry weight basis, the application rates for the compost, sawdust and sawdust/chicken litter were the equivalent of 12.8, 5.8 and 7.8 t/ha, respectively.

Tip cuttings of sweetpotato cv. Orleans were hand-planted on 3 December 2018 at a spacing of 6 plants/m row. The cuttings were planted with 3–5 nodes in the soil or amendment, with every effort made to ensure that they were placed in the middle of the V-furrow. Beds were irrigated with trickle tape and when the crop was inspected on 23 January 2019 (51 days after planting) there were no obvious differences in above-ground biomass between treatments. All plants grew well, with shoots covering the surface of the bed and growing into the furrow between the beds.

To assess early effects of treatments on the number of nematodes that had invaded roots, two sets of root samples were collected 51 days after planting. The first set was obtained by digging directly under the plant with a trowel and retrieving a sample of roots growing in the furrow where the amendments had been placed. Roots from the non-amended control were obtained from the same position. A second set of samples was obtained in the same way by removing roots from the area 5–10 cm outside the amended zone. Thus, a total of 96 root samples were collected: 48 from within the V-furrow and 48 from the soil outside the furrow.

Each root sample was washed thoroughly to remove any soil particles and blotted with towels and paper tissues to remove excess moisture. Any roots greater than 1 mm in diameter were removed because females and egg masses in swollen roots are embedded within tissue, and so they are difficult to locate and quantify. The fine roots remaining were weighed, with the fresh weight of most of the samples ranging from 0.2 to 0.8 g (mean 0.48 g). Each sample was then cut into segments with scissors, the segments were placed in water in a Petri dish and the roots observed under a dissecting microscope. Galls caused by root-knot nematode were readily apparent at a magnification of 10X and the number of galls in each root system was counted.

The sweetpotato crop was harvested 28 weeks after planting (14 June 2019). Storage roots were dug by hand from a 1 m section in the middle of each plot and yield data, together with levels of nematode damage, were obtained using procedures described previously.

As the results at harvest indicated that some treatments had reduced root-knot nematode populations and the levels of nematode damage, a range of tests were undertaken to categorise the soil biology and assess the nematode-suppressive capacity of the amended soils. The soil used for these tests was a 1 L sample collected from each plot immediately after the sweetpotatoes were harvested. Although the V-furrow zone was sampled, the disturbance that had occurred during the harvesting process meant that samples from amended plots were a mixture of amendment and soil.

Nematodes were extracted using methods described previously, but because the bulk density of the soil amended with sawdust was about 8% lower than other soils, nematode counts are reported as nematodes/200 mL soil. A detailed nematode community analysis was also undertaken by fixing the nematode suspension in 4% formaldehyde and identifying about 200 nematodes to genus or family level.

The nematode-suppressive capacity of the soils was assessed by placing 100 mL of soil into a screw-capped vial (volume 250 mL) and then adding Radopholus similis, a nematode that was not naturally present in the soil. Nematodes were cultured on sterile carrot tissue (O’Bannon and Taylor 1968; Moody et al. 1973) and 800 nematodes (a mixture of adults, juveniles and eggs) were added to each vial. A sample of soil from the trial site heated at 80 °C for an hour was used as a non-suppressive standard, and it was also inoculated with R. similis. After incubating the bioassay samples for 8 days at ambient temperatures (20–28 °C), nematodes were extracted using the tray method described previously and numbers of R. similis and free-living nematodes were counted.

Four groups of nematophagous organisms that may have been parasitising or preying on nematodes were assessed in the following ways. Nematode-trapping fungi were isolated using the sprinkle-plate method (Gray 1987) in which 1 g of soil was added to a Petri dish containing ¼-strength corn meal agar. Plates were checked after 10, 15 and 20 days for the presence of various fungi. The occurrence of fungi capable of parasitising root-knot nematode eggs was determined by collecting fine roots from the V-furrow zone after harvest, checking egg masses under a microscope and if parasitised eggs were observed, isolating the fungus on ¼-strength corn meal agar. Microarthropods were extracted by placing 350 mL of soil in a Tullgren funnel for 7 days (Walter and Kranz 2009) and capturing the animals in vials containing 70% ethanol. The presence of Pasteuria penetrans was assessed by checking second-stage juveniles of root-knot nematode at a magnification of 400X and noting whether they were encumbered with endospores.

Statistical analysis

The alternative farming system was assessed by measuring the nematode population, observing the performance of the sweetpotato crop, assessing various parameters in four 50 m plots and presenting the data as means ± standard error.

The V-furrow experiment included eight treatments resulting from the factorial combination of two bed formations (early and late) and four amendments (nil, sawdust, compost and sawdust/chicken litter). The experimental design was a split-plot with four main plots to which the two bed formation treatments were randomly assigned. Within each main plot, there were three replicate blocks containing four sub-plots each. The amendment treatments were randomly assigned to each of the four sub-plots within each replicate block. The experiment had 48 plots (four main plots x three replicate blocks x four sub-plots).

All analyses were done by analysis of variance (ANOVA) using GenStat 19, VSN International. Three variables needed a log transformation to stabilise the residual variance. Examination of the residual plots from the ANOVA for each variable did not indicate severe deviations from the assumptions of residual normality and homogeneity of variance. Mean comparisons were performed using the Least Significant Difference (LSD). In the suppression assay, one-tailed t-tests were used to determine whether the number of R. similis recovered from each of the amendments was significantly lower than the mean number recovered from composite samples of heated soil. A 5% significance level was used for all tests.

Results

Assessment of an alternative sweetpotato farming system

Although the cover crops were planted on beds that were formed 10-months earlier than is normal practice, the forage sorghum and oats grew well and both produced large amounts of biomass. As the aboveground residues produced by these crops were retained as mulch, the soil surface was covered with a thick layer of leaves and stems for most of the cover cropping period. This mulch was still present on the undisturbed beds prior to planting sweetpotatoes (Fig. 3a) and did not cause any problems when the crop was planted with minimal soil disturbance. The sweetpotatoes established well and grew in much the same manner as the rows which had been formed conventionally 2 weeks before planting (Fig. 3b, c). The crop went on to produce an acceptable yield (13.9 ± 1.9 kg of swollen roots/m row), which on a bed spacing of 1.5 m is equivalent to 92.8 ± 12.8 t/ha.

Fig. 3
figure 3

Two early-formed, organically-amended, non-disturbed beds are marked with arrows. The other beds were formed conventionally 2 weeks before planting. a. Immediately prior to planting. b. Four days later, after the beds had been strip-tilled with a tine and sweetpotatoes planted. c. Sweetpotato plants 7 weeks after planting

Full size image

Numbers of free-living nematodes in fields used for vegetable production in Queensland and northern NSW generally range from 200 to 2000 nematodes/200 g soil (GR Stirling, unpublished data) and as the site chosen for the study had 1934 ± 93 nematodes/ 200 g soil, the initial count was within the expected range (Table 1). Numbers increased markedly when the sawdust/chicken litter amendment was added to soil, remained very high throughout the cover cropping period, and were still relatively high when the sweetpotato crop was harvested (Table 1). Analyses of the free-living nematode community in soil collected at harvest showed that this was due to the presence of large numbers of bacterivores. Consequently, the Nematode Channel Ratio was 0.90 ± 0.03, indicating that the decomposition channel in the detritus food web was dominated by bacteria. Omnivores comprised only 10% of the nematode community and predatory nematodes (Order Mononchida and family Tripylidae) were not seen.

Table 1 Temporal changes in populations of root-knot nematode (Meloiodgyne incognita) and free-living nematodes in a field where an organic amendment was applied, beds were prepared 10 months prior to planting sweetpotato, two cover crops were grown on the beds and their residues were retained on the soil surface as mulch, sweetpotato was planted using strip tillage, and the crop was grown for 31 weeks
Full size table

Root-knot nematode (Meloidogyne incognita) was the dominant plant parasite in the field and by the time the beds were formed and forage sorghum was planted, numbers in the amended area were much lower than 6 weeks previously (Table 1). In contrast, there were 470 ± 56 root-knot nematodes/200 g soil in five samples of non-amended soil collected about 10 m from the area amended with sawdust and chicken litter. Numbers of root-knot nematode remained low in the amended, early-formed beds throughout the cover cropping period, and although a few second-stage juveniles (1–10 nematodes/plant) were extracted from 9 of the 16 oat root systems that were checked for the presence of the nematode, the population was very low when sweetpotatoes were planted.

The root-knot nematode population increased on the sweetpotatoes and by the time the crop matured there had been a large increase in nematode numbers (Table 1). Numbers were much higher 25 weeks after planting than when the sweetpotatoes were harvested, but this was probably due to differences in the sampling process. The in-crop samples consisted of root-associated soil whereas the soil collected after harvest was diluted with some bulk soil during the harvesting process. Given the very high nematode population at harvest, it was not surprising that the crop suffered severe damage from root-knot nematode. About 70% of the roots were unmarketable as they had damage ratings of 3 or 4.

The soil chemical and biochemical results mirrored the counts of free-living nematodes. Total soil C and soil N, permanganate-oxidisable C, and CO2 respiration assessed using the CO2-burst procedure all increased markedly in the area where the new farming system was tested, with these effects being apparent 7 weeks after the sweetpotatoes were planted, and also when the crop was approaching harvest (Table 2).

Table 2 Effects of an alternative sweetpotato farming system on total soil carbon and nitrogen, permanganate oxidisable carbon and microbial activity (assessed as CO2 respiration) prior to modifying the farming system and 13 and 17 months later, during the growth of a sweetpotato crop
Full size table

Field trial with organic amendments in a V-furrow

Nematode counts from the soil samples collected prior to applying the amendments showed that root-knot nematodes were distributed reasonably evenly across the trial site. The nematode population was similar in the two bed formation treatments, with counts in the beds formed early and late being 35 ± 10 and 40 ± 13 root-knot nematodes/200 g soil, respectively.

Analysis of the fresh weights of the root samples collected 51 days after planting and checked for galling showed that they were not significantly affected by bed formation time, amendment treatment or sampling position, indicating that similar-sized samples were assessed, regardless of treatment or the position where they were collected (data not shown).

Bed formation time did not affect the level of galling on roots collected from the V-furrow but amending the soil with organic matter had major effects. Significantly fewer galls were observed on roots collected from amended furrows than the non-amended control, with sawdust/chicken litter having the lowest gall counts (Table 3). In the area outside the amended zone, compost and sawdust did not reduce the level of galling relative to the non-amended control whereas sawdust/chicken litter significantly reduced the number of galls.

Table 3 Effect of organic amendments applied to a V-furrow in the centre of the bed on the level of galling caused by root-knot nematode (Meloidogyne incognita) on roots growing within or outside the amended zone 51 days after sweetpotatoes were planted
Full size table

When the crop was harvested after 28 weeks, bed formation time did not interact with amendment for any of the analysed variables whereas the amendment effect was always significant. The root-knot nematode population was very high in the non-amended control but was significantly lower in all amendment treatments (Table 4). The damage ratings obtained by assessing each storage root largely mirrored that result, as the rating was significantly lower in the sawdust-based amendments than the non-amended control. Although all amendments increased total yield, sawdust and sawdust/chicken litter were the only amendments to significantly increase marketable yield (Table 4).

Table 4 Effect of organic amendments applied to a V-furrow in the centre of the bed on populations of root-knot nematode (Meloidogyne incognita, RKN), the severity of nematode damage, and total and marketable yields of a sweetpotato crop harvested 28 weeks after planting
Full size table

A detailed assessment of the free-living nematodes extracted from soil collected after harvest showed that regardless of the treatment, more than 90% of the nematodes were in six groups. The Rhabditidae and Cephalobidae were the main bacterivores, Aphelenchoides was the predominant fungivore, two groups of plant associates were always present (Dorylaimellus and the fine-tailed Tylenchidae), while the omnivores (order Dorylaimida) were represented by Eudorylaimus, Aporcellaimellus and the Discolaimidae. Interestingly, there were very few predatory nematodes, as mononchids were absent and members of the family Tripylidae were only observed occasionally.

Compost had no impact on the free-living nematode community, as its composition was always similar to the non-amended control. In contrast, sawdust had major effects, as the proportion of Aphelenchoides and Dorylaimellus was significantly higher and the proportion of omnivores significantly lower than all other treatments. Sawdust-amended soil also had a much lower Nematode Channel Ratio than the other treatments (0.61 compared with 0.77, 0.80 and 0.78 for nil, compost and sawdust/chicken litter, respectively). The main effect of sawdust/chicken litter was to increase the proportion of rhabditids in the nematode community to about 25%, whereas it was only about 5% in the other amended soils and the untreated control.

In the suppression assay in which 800 R. similis were added to screw-capped vials containing 100 mL soil, 534 ± 16 nematodes were recovered from the heated soil. Numbers were significantly lower in non-amended soil from the field, and were even lower in soils to which amendments had been added (Table 5). Sawdust had the greatest impact but sawdust/chicken litter also markedly reduced the number of R. similis recovered. They were also the treatments with the highest populations of free-living nematodes (Table 5).

Table 5 Numbers of Radopholus similis and free-living nematodes recovered in a suppression assay 8 days after 100 mL soil samples collected from plots where various organic amendments had been applied were inoculated with 800 R. similis
Full size table

When the soils were checked for nematode-trapping fungi, three fungi that produce three-dimensional network traps were observed: Arthrobotrys musiformis, A. thaumasia and A. oligospora. One or two of these fungi were observed in almost all plates, except that they were not observed as frequently in soils that had been amended with sawdust/chicken litter (data not shown).

Counts of microarthropods showed that numbers were much higher in soil amended with sawdust/chicken litter (405 ± 89 animals/350 mL soil) than in the other treatments, where the mean number extracted ranged from 50 to 66 animals/350 mL soil. The bulb mite (Rhizoglyphus robini) was by far the most common mite in all treatments (often more than 80% of the mite population), with most of the remainder being Oribatida and Astigmata. There were relatively few Mesostigmata (a group which prey on nematodes) and they were most commonly observed in sawdust-amended soil, sometimes comprising 3–4% of the mite community. The mesostigmatids identified belonged to the family Laelapidae and the genera Protogamasellus and Sejus.

When second-stage juveniles of root-knot nematode were checked for the presence of Pasteuria endospores, spore-encumbered nematodes were observed in all treatments. About 10% of the nematodes had some spores attached (usually 1–4 spores/nematode) but occasional nematodes were encumbered with more than 20 spores. There were no obvious differences between treatments (data not shown).

Observations of Meloidogyne egg masses retrieved from fine roots collected after harvest showed that in the non-amended control and all amended soils, some eggs were parasitised by fungi. In most cases, fungi were observed in 30–50% of the egg masses and when they were present, large numbers of eggs were parasitised. Two fungi were isolated from these egg masses: Pochonia chlamydosporia and Purpureocillium lilacinum.

Discussion

An alternative sweetpotato farming system

Our on-farm test of an alternative sweetpotato farming system clearly showed that it is possible to replace the tillage-dominated system currently used by Australian growers. The test system had five components (early bed formation, an organic amendment, cover crops suitable for summer and winter, residues from the cover crops retained as mulch, and strip-till planting) and when these practices were used together, the crop established well and produced yields comparable to those achieved with the conventional farming system.

Many years of research on a range of crops provides evidence that soil health will improve and the system will be more sustainable when all the above practices are integrated into the farming system. Minimising tillage is perhaps the most important component because it helps retain some of the organic carbon that would otherwise be lost to the atmosphere under more aggressive tillage regimes (Magdoff and Weil 2004). It also lessens the detrimental effects of tillage on the fungi, microarthropods, earthworms and other soil organisms that provide many important ecosystem services (Wardle 1995; Kabir 2005; Lehman et al. 2015; Gupta et al. 2019). Although our system does not eliminate tillage, it reduces the number of tillage events and their intensity. Also, beds are formed many months before planting and so the soil biology has a chance to partially recover from the disturbance that inevitably occurs when the crop is harvested.

Including cover crops in the farming system is another important component of sustainable agriculture (Blanco-Canqui et al. 2015) and this was achieved in our system by growing forage sorghum in summer and oats in winter. Although the inclusion of two cover crop species rather than one may have increased soil biodiversity to some extent, multi-species cover cropping is likely to have an even greater impact and certainly warrants further testing, provided the crops are sown in ways that minimise soil disturbance. Because oats grows much better than forage sorghum in winter, perhaps the greatest benefit obtained from adding a second cover crop was a substantial increase in the amount of dry matter produced.

Oats cv. Saia was chosen as the winter cover crop, but its susceptibility to M. incognita is a shortcoming (Stirling et al. 1996). Although our observations suggested that only limited nematode multiplication occurred, it would have been better to terminate the oats earlier. Meloidogyne incognita requires about 410 day-degrees to complete its life cycle (Dávila-Negrón and Dickson 2013; Hay and Stirling 2014) and temperature data from the nearest site where records were available (Coolangatta airport) indicated that egg production would have commenced in early September. Thus, if the oats had been sprayed with herbicides 3 weeks earlier, nematode multiplication could have been prevented. Other alternatives would be to grow oats cv. Algerian, which is relatively resistant to M. incognita (Stirling et al. 1996), or undertake a screening program to find resistant cover crops that grow well in winter.

Retention of cover crop residues as mulch is important from a soil health perspective because it reduces wind and water erosion, helps retain soil moisture, dampens temperature fluctuations and provides carbon inputs that nourish the soil biological community. It is also beneficial from a nematode management perspective because research in sugarcane has shown that the blanket of mulch which remains after the crop is harvested enhances the soil’s suppressiveness to plant-parasitic nematodes (Stirling et al. 2011a, b). Others have shown that sweetpotato can be grown successfully using cover crop residues as mulch (Jackson and Harrison 2008) and mulching also provides worthwhile levels of weed control (Nwosisi et al. 2017).

An organic amendment was integrated into the new farming system because previous work had shown that when a mixture of sawdust and poultry manure was applied at 200 t/ha, the amended soil was suppressive to root-knot nematode for 2 years (Stirling et al. 2012). Initial results with the sawdust/chicken litter used in this study were encouraging because root-knot nematode numbers declined by about 99% in the first 6 weeks. This effect was almost certainly due to ammonia, as it is nematicidal at high concentrations (Eno et al. 1955). Chicken litter contains relatively large amounts of ammoniacal nitrogen (Weidemann 2015) and many studies have shown that the ammonia released from animal manures with a low C/N ratio provides immediate and high levels of nematode control (e.g. Mian and Rodriguez-Kabana 1982; Rodriguez-Kábana 1986; Rodriguez-Kábana et al. 1987). However, we obtained no evidence to indicate that the amendment enhanced the suppressive services provided by the soil biota, as results obtained later in the study indicated that root-knot nematode populations remained at much the same level for the next 9 months and then increased exponentially when sweetpotato was planted.

In looking for reasons why the sawdust/chicken litter amendment failed to enhance biological suppressiveness and provide long-term control of root-knot nematode, two issues stand out. First, the sweetpotatoes grew for 31 weeks before being harvested and given the reproductive potential of M. incognita and the number of degree-days accumulated, the nematode would have completed five or possibly six life cycles in that time. In such a situation, natural enemies are unlikely to provide the level of control required. Second, observations of the free-living nematode community at harvest showed that the Nematode Channel Ratio was 0.9, indicating that the soil microbial community was dominated by bacteria rather than fungi. Fungal parasites and predators are perhaps the most important natural enemies of nematodes (Stirling 2014) and although sawdust was included in the amendment to stimulate these fungi, this did not occur. Presumably the chicken litter added with the sawdust had a negative impact on the fungal community, as nitrogen inputs are known to be detrimental to many fungal taxa, including the nematophagous genus Dactylella (Paungfoo-Lonhienne et al. 2015). As previous work has shown that root-knot nematode is suppressed by sawdust amendments when urea and poultry manure are used as nitrogen sources (Vawdrey and Stirling 1997; Stirling et al. 2012), future research should focus on determining the rate and type of nitrogen that should be added to sawdust to achieve a diverse fungal community capable of reducing populations of root-knot nematode.

Application of organic amendments in a V-furrow

The trial in which sweetpotatoes were planted into furrows filled with three different amendments produced much more promising results from a nematode control perspective. When roots were collected from the centre of the bed about 7 weeks after planting, the number of galls produced by root-knot nematode was much lower where organic amendments had been applied than in non-amended soil. Sawdust/chicken litter had a major effect, reducing the number of galls within and outside the amended zone. Compost and sawdust reduced galling by 71% and 56%, respectively, but this effect was only observed on roots growing in the amended zone.

Although final root-knot nematode populations were high, the results obtained at harvest were also encouraging, as both sawdust and sawdust/chicken litter increased marketable yield by 29% and reduced nematode populations by 49% and 39%, respectively. The marketable yield in both treatments was the equivalent of 92.7 t/ha, a yield that would be acceptable to most Australian growers, and much higher than is normally obtained in other countries. Compost increased total yield but as it did not reduce the damage rating or increase marketable yield, its effects mirrored the results obtained with composts in other studies (Thoden et al. 2011).

At this stage, it is only possible to speculate on the reasons why two of the organic amendments reduced galling in the amended zone, but the early effects of sawdust/chicken litter were probably nitrogen-related, as ammonia released from the chicken litter would have been nematicidal (see earlier discussion). It is also possible that the chicken litter was detrimental to early root growth, as amendments containing chicken litter caused an initial suppression of plant growth in tomato and cotton (Kaplan and Noe 1993; Riegel and Noe 2000), while excess nitrogen is known to reduce swollen root formation in sweetpotato (Taranet et al. 2017). Thus, the lower levels of galling may have been due to fewer roots being available to the nematode. Sawdust on its own was also an effective amendment, but because it proved difficult to wet, the soil remained relatively dry for the first week or so. Thus, some of its early effects may have been due to poor root growth, or to moisture limitations that prevented second-stage juveniles from moving to and invading roots.

The results of numerous studies over many years have shown that populations of nematophagous fungi, predatory mites and predatory nematodes usually increase when soil is amended with organic matter (Stirling 2014). We obtained some evidence that this occurred with sawdust and sawdust/chicken litter, suggesting that natural enemies contributed to the reduction in root-knot nematode populations observed with these amendments. Evidence from the suppression test supports that contention because fewer R. similis were recovered from amended than non-amended soil. The presence of high numbers of free-living nematodes in the amended soils also suggests that nematophagous activity would have been occurring, as these nematodes provide a food source for parasites and predators.

Conclusions

Our on-farm test of an alternative sweetpotato farming system showed that the crop established well and produced an acceptable yield when the soil was amended with a mixture of sawdust and chicken litter, beds were formed 10 months prior to planting, successive cover crops were grown on those beds, the residues produced by those crops were retained on the soil surface as mulch, and sweetpotatoes were planted using strip tillage. In addition, nematode counts taken 6 and 47 weeks after the organic amendment was applied showed that root-knot nematode populations had been reduced to very low levels. Measurements taken in the following sweetpotato crop also showed that there was an improvement in several soil health parameters (e.g. total and permanganate oxidisable C, microbial activity and numbers of free-living nematodes). Nevertheless, the sweetpotatoes suffered severe nematode damage, presumably because nitrogen inputs from the chicken litter produced a detritus food web dominated by bacteria rather than fungi, and so the soil lost its nematode-suppressive capacity.

The results of an experiment carried out in the same field at the same time were more encouraging from a nematode management perspective because organic amendments were effective when placed in a furrow in the centre of the bed so that the swollen roots were surrounded by the amendment as they developed. Sawdust and a mixture of sawdust/chicken litter both reduced root-knot nematode populations and the severity of nematode damage, and increased marketable yield by 29%. In addition, a diverse range of natural enemies was detected in soil collected from the amended furrow at harvest. When R. similis was added to the amended soils, numbers were markedly reduced, indicating the presence of a soil biological community capable of suppressing plant-parasitic nematodes. Further evidence that sawdust-based amendments are biologically suppressive can be found in a follow-on paper (Stirling 2020).

The tillage-dominated farming system currently used by Australian sweetpotato growers is detrimental to soil health, but the above results demonstrate that it should be possible to replace it with a more sustainable alternative. The process involved has been outlined elsewhere (Stirling et al. 2016) but basic practices such as controlled traffic, early bed formation, cover cropping, minimum tillage and organic amendments will be key components of any new system. Although it is a challenging task to modify a productive and well-established farming system, the first tentative steps have now been taken. Provided land managers are willing to assess the above practices in on-farm trials, learn from the results, and then integrate some or all of these practices into their farming system, the end result is likely to be healthier soils and more sustainable farm businesses.