Holding it together soils for sustainable vegetable production

Transcription

1 Holding it together soils for sustainable vegetable production Johnstone PR, Wallace DR, Arnold N, Bloomer D July 2011 A report prepared for Horticulture New Zealand Johnstone PR, Wallace DF, Arnold N Plant & Food Research, Hawke s Bay Bloomer D LandWISE, Hawke s Bay SPTS No. 5756

2 DISCLAIMER Unless agreed otherwise, The New Zealand Institute for Plant & Food Research Limited does not give any prediction, warranty or assurance in relation to the accuracy of or fitness for any particular use or application of, any information or scientific or other result contained in this report. Neither Plant & Food Research nor any of its employees shall be liable for any cost (including legal costs), claim, liability, loss, damage, injury or the like, which may be suffered or incurred as a direct or indirect result of the reliance by any person on any information contained in this report. This report has been prepared by The New Zealand Institute for Plant & Food Research Limited (Plant & Food Research), which has its Head Office at 120 Mt Albert Rd, Mt Albert, Auckland. This report has been approved by: Paul Johnstone Scientist - Land Management Date: 29 July 2011 Steve Thomas (Acting) Science Group Leader - Soil, Water and Environment Date: 29 July 2011

3 Contents Executive summary i 1 Introduction 1 2 Overview of methods Background Statistical analyses 2 3 Cover crops Selecting cover crop options for vegetable rotations 3 Case study 1: Biofumigant cover crops 3 Case study 2: Types of cover crops Managing crop residues Case study 3: Management of cereal straw Using cover crops to recycle nutrients Case study 4: Maize to recover deep nutrients Surface runoff Reducing the flow of water across the soil surface Case study 5: Wheel track diking Using secondary crops to reduce soil losses 24 Case study 6: Crop stabilisers in the wheel track 25 Case study 7: Crop stabilisers across the paddock Using chemicals to stabilise the soil Case study 8: Chemical stabilisers across the paddock Cultivation and soil compaction Controlled traffic farming approaches 32 Case study 9: Controlled traffic farming 32 Case study 10: Permanent beds Designated traffic zones to limit compaction damage 41 Case study 11: Designated traffic zones 42 6 Good Agricultural Practice (GAP) template 43 7 Acknowledgments 44

4

5 Executive summary Holding it together soils for sustainable vegetable production Johnstone et al, July 2011, SPTS No New Zealand s vegetable industry relies on a wide variety of essential services provided by the soil. These services can be directly affected by a number of cultural management practices, both in the short and longer term. There is a constant search for opportunities to increase the sustainability of intensive production practices that are currently used, including approaches that minimise potential environmental impacts, maintain or grow productivity, and protect market access. To help address this challenge, the vegetable industry, its partners, several regional councils and MAF-Sustainable Farming Fund supported a new initiative called Holding it Together (HIT). The overall aim of this project was to demonstrate how various soil management practices can be used to address existing environmental and production challenges. In the first year of the project the most pressing challenges were identified during discussions with industry leaders and land management officers. These were broadly grouped into the following themes: cover crops, surface runoff, and soil cultivation and compaction. This report summarises demonstrations from the final 2 years of this project ( , ) within these themes. Theme I: Cover crops Selecting cover crops: - There are a wide variety of cover crops available, many of which offer different benefits. Identifying the intended use is therefore a critical step in ensuring cover crop use is profitable. To get the most from cover crops they should be regularly integrated into the rotation. Incorporation and traffic practices should also be designed to limit any unnecessary cultivation and compaction. The cost of cover crops is often offset by direct and measurable benefits such as the return of nutrients to the soil and reduced herbicide use. Other benefits such as improved soil physical quality may be harder to value, but are clearly linked to the long term productivity and resilience of the soil. Managing cereal residues: - In many cases exporting straw residues is more profitable than retaining the residue. However, if there is no plan to return organic matter back to the soil over the long-term (either as crop residues or other break and green crops) soil organic matter content will typically decrease. Recycling nutrients with cover crops: - Maize and other deep rooted break crops can recycle nutrients from depths that are not typically explored by many vegetable crops. This provides an opportunity to increase nutrient use efficiency and reduce fertiliser costs. It also minimise the potential loss of mobile nutrients to the environment. Theme II: Surface runoff Reducing the flow of water across soil surfaces: - Reducing surface runoff and soil erosion from cultivated paddocks starts by preventing water from entering the paddock from adjacent areas. This can be achieved through the use of interception drains The New Zealand Institute for Plant & Food Research Limited (2011) Page i

6 and bunds. However, control measures may also be necessary to prevent in-field water movement. Wheel track diking is one approach that can reduce surface runoff across a wide range of conditions. Diking equipment is comparatively cheap and can be attached to existing implements. The frequency of diking depends on the conditions in which it is used. In some cases one pass per season may be enough, whereas in others several passes may be necessary, between rainfall or irrigation events. Using secondary crops to reduce soil losses: - Secondary crops can be used to stabilise soils and reduce erosion during periods of high risk. Quick establishment of these crops across the whole bed minimises soil losses. Using chemical stabilisers to reduce soil losses: - Chemical stabilisers showed potential to settle out sediments in laboratory conditions. Further testing is still necessary to confirm the field performance of these products. Theme III: Cultivation and soil compaction Using controlled traffic farming (CTF) and permanent beds to reduce compaction and cultivation: - CTF and permanent bed approaches provide growers with an opportunity to limit compaction damage and reduce the amount of cultivation necessary between crop sequences. This can have a major impact on fuel use efficiency, as well as on soil physical quality. Using designated traffic zones: - Designated traffic lanes provide an effective means of confining compaction damage, especially if the crop is being harvested when the soil is wet. Because these areas are routinely used during the season they also tend to be more consolidated, which can actually reduce delays accessing fields following heavy rainfall events. These traffic areas can be retained from crop to crop, reducing the need to cultivate heavily compacted or damaged areas. Alongside these practical field demonstrations a Good Agricultural Practice (GAP) template was developed that allows growers to demonstrate compliance with appropriate regulatory and market based requirements related to the management of their soils. The document is available online at: For further information please contact: Paul Johnstone The New Zealand Institute for Plant & Food Research Limited Plant & Food Research Hawke s Bay Private Bag 1401 Havelock North 4157 NEW ZEALAND Tel: Fax: Paul.Johnstone@plantandfood.co.nz DDI: The New Zealand Institute for Plant & Food Research Limited (2011) Page ii

7 1 Introduction New Zealand s vegetable industry relies on a wide variety of essential services provided by the soil. These include its ability to store and supply water, retain and release nutrients, provide aeration, and stabilise crops during growth. These services can be directly affected by a number of cultural practices used by growers, both in the short and longer term. There is a constant search for opportunities to increase the sustainability of intensive production practices that are currently used. This includes opportunities to minimise potential environmental impacts, maintain or grow productivity, and protect market access through the adoption of good agricultural practices. To help address this challenge, in 2008 the vegetable industry, its partners, several regional councils and MAF-SFF funded a new 3-year initiative called Holding it Together (HIT). The overall aim of the project was to demonstrate how various soil management practices can be used to address existing environmental and production challenges. The primary regions that were targeted by the HIT project were Horowhenua, Pukekohe and Hawke s Bay; these areas encompass a significant amount of the land area utilised for outdoor vegetable production in New Zealand. The findings from the project are however equally applicable in other cropping regions. In the first year of the project the most pressing challenges were identified during discussions with industry leaders and land management officers in each region. These were broadly grouped into the following themes (though many were clearly inter-related): cover crops, surface runoff, and soil cultivation and compaction. A key feature of the project was that researchers and technologists worked alongside growers to trial established and innovative practices on commercial properties. In some cases this involved measuring the impacts of changes in management, whereas in others it involved working through the practical issues and implications associated with a particular technology or technique. In addition to these practical demonstrations a Good Agricultural Practice (GAP) template was developed that allows growers to demonstrate compliance with appropriate regulatory and market based requirements related to the management of their soils. This report summarises a series of field demonstrations from the final 2 years of this project ( , ); observations from the first year ( ) were summarised in an earlier report that is available online ( The New Zealand Institute for Plant & Food Research Limited (2011) Page 1

8 2 Overview of methods 2.1 Background The emphasis of field trials conducted within the HIT project was on demonstrating a variety of management practices that can improve the overall quality, productivity, and resilience of soils used for vegetable cropping. To achieve this, basic experimental designs were often favoured over more complex approaches. This was necessary because all trials occurred on commercial properties and needed to be implemented at a scale that was both practical and informative to the collaborating growers. The trade-off was that a number of trials had a limited number of replicates. In these cases every effort was made to eliminate or address potential sources of variation that are typically handled by more complex trial designs. The type and frequency of measurements made at each site varied considerably depending on the purpose of the trial and the interest of the collaborating grower. These are summarised in brief for each case study. 2.2 Statistical analyses The main statistical approaches used to analyse data from these field demonstrations were single-factor ANOVAs or t-tests. In some instances low replication limited the statistical power that was possible for a given analysis. Significance values were recorded where P < 0.05; P values between 0.05 and 0.10 were considered weakly significant and should be interpreted with appropriate caution. P values above 0.10 were denoted as not significant (ns). Least significant difference (LSD) values were calculated to separate treatment means (α = 0.05) when using ANOVA approaches, and represent the smallest difference necessary between two means for a statistically significant test. The New Zealand Institute for Plant & Food Research Limited (2011) Page 2

9 3 Cover crops Cover crops are used widely in the New Zealand vegetable industry. Some of the key benefits that can result from their use include: higher soil organic matter content, improved nutrient cycling, less soil erosion, quicker drainage, disruption of pest and disease cycles, fewer weeds, reduced compaction, and greater biodiversity. Importantly, some of these benefits from cover crops are realised quickly in a matter of months, whereas others can take years to develop. One of the most common challenges associated with the use of cover crops is how they can be integrated profitably into cropping systems. This often influences the frequency with which cover crops are used and also the variety of cover crops that are selected, factors that have a major influence on the type and degree of benefit. To help growers with these decisions a number of demonstrations were established to highlight a range of considerations that are important when selecting and managing cover crops. Key areas of grower interest centred on selecting the right cover crop option, retaining and recycling nutrients and managing crop residues. Figure 1: Cover crops can be used to improve profitability and productivity within vegetable cropping systems. 3.1 Selecting cover crop options for vegetable rotations A wide variety of cover crop options are available on the market. In many settings annual grasses and cereals have been favoured, often because they are quick to establish, easy to manage and hardy across a wide range of conditions. However, many other cover crop options that provide additional soil conditioning benefits are available. Being clear on the purpose of a cover crop is central to maximising the potential benefit resulting from their use. Case study 1: Biofumigant cover crops Approach In 2009 a demonstration trial was set up in Horowhenua to compare the performance of a small number of cover crop options to the grower s standard fallow approach over the winter. Soil texture was clay loam, and the paddock had been intensively cropped for about 30 years. Cover crop options tested included annual ryegrass and two Caliente mustard blends (Nemat and ISCI 99); both mustards have been specifically bred to have a fumigant-like effect when incorporated back into the soil (a benefit of interest to the collaborating grower). During autumn the field was conventionally cultivated to prepare a suitable seed bed. The cover crops were subsequently sown in late June using a small plot drill. Experimental design was a randomised complete block, with four replicates of each cover crop. The size of each individual The New Zealand Institute for Plant & Food Research Limited (2011) Page 3

10 plot was approximately 25 m 2 (7 m x 3.6 m). The cover crops received no additional management (herbicides, fertiliser, or fungicides), which reflected the typical grower approach. Table 1: What were the key cover crop characteristics, sowing rates and seed costs? Cover crop Key crop characteristics 1 Sowing rate (kg seed/ha) Moata ryegrass GC, G, S 30 Nemat mustard B, GC 8 Total seed cost (per ha) 2 $114 (@ $3.80/kg) $220 (@ $27.50/kg) $213 ISCI 99 mustard B, GC 12 (@ $17.75/kg) 1 GC = green crop, G = grazing, S = silage, B = biofumigant. 2 Indicative cost only (for up-to-date figures contact a local seed supplier). The first stage of this trial concluded in late September (approximately 3 months after cover crops were sown). At that time, biomass yields were measured in 1 m 2 quadrats taken from every plot and a subsample analysed for total N concentration. An estimate of root biomass in the top 15 cm of soil was also made by collecting three cylindrical cores (8 cm diameter) from each plot and extracting roots using wet sieving. Soil mineral N content was measured on a composite sample taken from each treatment (two cores per replicate). Sampling depths were 0 30 cm and cm, reflecting the maximum rooting depth of many vegetable crops. Soil bulk density and aggregate stability were also measured to assess the short term impact of winter cover crops on these indicators of soil physical condition. These measurements were based on a three core sample (0 15 cm) collected from every plot. Cover crop residues were then immediately incorporated using several rotary hoe passes. This was necessary because the residues had not been mulched first. Figure 2: Early growth at the site in mid August (left) and at harvest in late September (right). In early November lettuce seedlings were transplanted into beds that had been formed within the existing trial design. There were three rows of lettuce per bed. The lettuce crop received no additional fertiliser or irrigation due to the small trial area; however, background soil fertility was already high in the field (Olsen P = 146 mg/l, MAF quick test K = 30 units, mineral N ~125 kg N/ha). Lettuce yields were measured once the crop reached commercial maturity (late December). The harvested area in each plot was 2.5 m x 3 lettuce rows wide, from which the number of marketable heads and their fresh yield (total head weight, as well as marketable head weight) was recorded. The number of heads affected with lettuce leaf drop (caused by Sclerotinia minor) was also recorded to assess the control of this soil-borne disease resulting from the The New Zealand Institute for Plant & Food Research Limited (2011) Page 4

11 earlier incorporation of the biofumigant mustards. The total number of emerged weeds was also recorded using counts from small 1m 2 quadrats randomly positioned in each plot. Soil water movement was then measured in all plots using a single ring infiltration test. This test was designed to highlight relative differences in drainage between each prior cover crop approach, rather than to quantify absolute differences in the rate of water movement. There were four rings per plot, and each ring was inserted 4 cm into the soil ahead of testing (rings were 15 cm wide x 15 cm long). A constant amount of water was applied to each ring ahead of the test to ensure the starting soil moisture was similar for all plots. The time taken to drain 4 cm of water was then recorded for each core. The test was repeated five times in quick succession (i.e. a total amount of 20 cm of water was applied), and an average drainage time was calculated for each plot. Findings Table 2: What was the effect of cover crop selection on biomass and N dynamics immediately prior to incorporation? Cover crop Crop biomass (t DM/ha) Root biomass (t DM/ha, 0 15 cm) Crop N uptake (kg N/ha) 1 Soil mineral N (kg N/ha, 0 60 cm) N in the system (kg N/ha) 2 Fallow Ryegrass Nemat ISCI Statistical significance 3 ns (1.1) <0.001 (0.3) ns (24) Above-ground crop component only. 2 Calculated as the balance of N in the crop and soil (0-60 cm) at incorporation. 3 Significance test is based on a general ANOVA (n = 12); ns = not significantly different at P < Values in parentheses represent the least significant difference (LSD). 4 Because soil mineral N data were unreplicated they were not analysed statistically. General observations: There was no difference in above-ground biomass between the cover crops. Overall, biomass yields were low, reflecting both the late planting date in winter and the early incorporation date in spring. Ryegrass had significantly more root biomass in the top 15 cm than either mustard. This reflected the shallow sampling depth and the fact that mustards have a taproot. It is important to note that both mustards were grown for their biofumigant potential rather than prolific rooting and direct soil conditioning characteristics. All cover crops accumulated a considerable amount of N, which resulted in lower mineral N concentration in the soil at incorporation (29 45% less mineral N than the fallow approach). Soil mineral N is susceptible to leaching and denitrification (especially during the winter), so minimising soil N concentration by cover cropping is an effective means to limit such losses. Estimates of the summed mineral and crop N at incorporation showed that the fallow control had lost on average 68 kg N/ha compared to the other cover crop options. This likely reflected leaching of N below the sampling depth of 60 cm (beyond the rooting depth of many vegetable crops). The replacement cost of this N using urea fertiliser is approximately $110/ha (at $720/t of urea, assuming 100% efficiency of the applied The New Zealand Institute for Plant & Food Research Limited (2011) Page 5

12 fertiliser), which represented a significant component of the initial cost of the cover crop seed. Table 3: What was the effect of cover crop selection on soil physical characteristics in the top 15 cm immediately prior to incorporation? Cover crop Soil bulk density (g/cm 3 ) Aggregate stability (MWD, mm) 1 % of aggregates > 1 mm 2 Fallow Ryegrass Nemat ISCI Statistical significance 3 ns (0.1) 0.09 (0.11) 0.07 (5) 1 Represents the mean aggregates size after a standardised wet sieving test (an indicator of stability); MWD = mean weight diameter. 2 Represents the percent of aggregates that have a MWD of greater than 1 mm after the wet sieving test. 3 Significance test is based on a general ANOVA (n = 16); ns = not significantly different at P < Values in parentheses represent the least significant difference (LSD). General observations: There was no significant effect of cover crop selection on soil bulk density in the top 15 cm at incorporation. This observation was not surprising, as bulk density is a comparatively coarse indicator of potential soil conditioning effects over the duration of this trial. There was some effect of cover crop selection on soil aggregate stability. In particular, ryegrass resulted in a greater mean aggregate size and fraction of stable aggregates following the wet sieving process than the fallow control (a higher score is often an indication of improved physical structure). Both mustards resulted in intermediate outcomes for these indicators. These findings appeared consistent with the differences in root biomass that were observed between the cover crops options (i.e. highest in the ryegrass plots). In all cases MWD values were well below 1.5 mm, a threshold below which studies have shown a higher likelihood of poorer crop yields. This observation was apparently linked to the history of regular and intensive cultivation at this site. Table 4: What was the effect of previous cover crop selection on lettuce yield, disease incidence and weed populations in the next rotation? Previous cover crop Total number of heads (per ha) Marketable fresh mass (kg/head) Disease incidence (%) Weed population (per m 2 ) Fallow 39, Ryegrass 41, Nemat 36, ISCI 99 39, Statistical 3 ns (6000) 0.06 (0.166) ns (12) ns (8) significance 1 Marketable heads had outside wrapper leaves removed according to commercial practice. 2 Calculated as the percent of total heads in each plot that had lettuce leaf drop. 3 Significance test is based on a general ANOVA (n = 16); ns = not significantly different at P < Values in parentheses represent the least significant difference (LSD). General observations: Small differences in the total number of heads counted at harvest reflected variation in the mortality of transplanted lettuce rather than any effect of prior cover crop selection. The New Zealand Institute for Plant & Food Research Limited (2011) Page 6

13 Marketable fresh mass of individual lettuce heads was significantly higher in ISCI plots than in fallow or ryegrass plots. This appeared likely to reflect differences in N availability between treatments. For example, fallow plots lost a substantial amount of plant-available N during the winter (68 kg N/ha compared to the cover crops). While ryegrass had a similar N balance to the mustard plots when the cover crops were incorporated, it was also much slower to break down. These yield outcomes were clearly influenced by the fact that no N fertilisers were applied to the lettuce crop. While this is not standard practice, it did highlight opportunities for growers to reduce fertiliser use through the use of cover crops. There was no effect of mustard cover crops on disease incidence or weed severity in lettuce. The absence of any notable biofumigant effect in this trial was related to the relatively small amount of mustard biomass produced owing to the late planting (~3 t DM/ha) and the fact that recommended incorporation practices were not followed. Figure 3: Example of lettuce leaf drop observed at the trial site in December (left) and of the single ring infiltration approach used to assess relative differences in water drainage in the next rotation (right). Figure 4: What was the effect of previous cover crop selection on water infiltration rate in the next rotation? General observations: The time taken to drain 40 mm of water into the soil was variable across plots. Overall though, water infiltration appeared to be slowest in the fallow control and quickest in the ryegrass plots. This likely reflected the influence of cover crops rooting characteristics on soil conditioning. The New Zealand Institute for Plant & Food Research Limited (2011) Page 7

14 Case study 2: Types of cover crops Approach In 2010 a demonstration trial was set up in Horowhenua to enable growers to assess the relative the performance of a wider range of winter cover crop options side by side. Soil texture was silt loam, and the paddock had a history of regular vegetable production with few break crops. The site was prepared for planting using conventional cultivation practices. A total of 14 cover crops were subsequently sown in mid-may using a small plot drill. The cover crops tested included annual ryegrass, oats, triticale, blue lupin, chicory, Nemat mustard, ISCI 99 mustard, a triticale-peas mix, common vetch, buckwheat, wheat, a hairy vetch-oats mix, phacelia and clover. A summary of cover crop characteristics, sowing rates and seed costs is provided below. Two fallow controls were also included in the trial; one was kept chemically fallow (the grower standard), whereas in the other weeds were left unsprayed. Table 5: What were the key cover crop characteristics, sowing rates and seed costs? Cover crop Key crop characteristics 1 Sowing rate (kg seed/ha) Crusader ryegrass GC, G, S 25 Milton oats GC, G, S 120 Profit triticale GC, G, S 120 Blue lupin L, GC, G 100 Chicory GC, G 5 Nemat mustard B, GC 8 ISCI 99 mustard B, GC 12 Total seed cost (per ha) 2 $150 (@ $6.00/kg) $192 (@ $1.60/kg) $192 (@ $1.60/kg) $295 (@ $2.95/kg) $90 (@ $18.00/kg) $220 (@ $27.50/kg) $213 (@ $17.75/kg) Double Take triticale + Provider peas 3 GC, L 120 $180 (@ $1.50/kg) Common vetch GC, L 45 Buckwheat GC, P 40 Wheat GC, S 100 Hairy vetch + Milton oats 4 GC, L 100 Phacelia GC, P 3 Bolta clover L, GC, G 3 $63 (@ $1.40/kg) $76 (@ $1.90/kg) $165 (@ $1.65/kg) $170 (@ $1.70/kg) $42 (@ $14.00/kg) $36 (@ $12.00/kg) 1 GC = green crop, G = grazing, S = silage, L = legume, B = biofumigant, P = predator habitat. 2 Indicative cost only (for up-to-date figures contact a local seed supplier). 3 The mix comprised approximately 70% triticale and 30% peas on a per weight basis. 4 The mix comprised approximately 30% vetch and 70% oats on a per weight basis. Experimental design was a randomised complete block with four replicates of each cover crop and control option. The size of individual plots was approximately 17 m 2 (7 m x 2.4 m). The cover crops received no additional management (herbicides, fertiliser, or fungicides), which reflected the typical approach used. The exception to this was the chemical fallow plots, which were sprayed twice with glyphosate during the winter to control weeds. The New Zealand Institute for Plant & Food Research Limited (2011) Page 8

15 Figure 5: Trial site in mid July once the crops were well established (left). Signs describing key crop characteristics and details were put in all plots to allow local growers to observe crop performance during the winter (right). The trial concluded in early October (approximately 5 months after sowing), at which time biomass yields were measured in 1 m 2 quadrats taken from every plot. The proportion of weed to cover crop biomass was also recorded. Total N concentration of each cover crop was determined on a composite sample taken across replicates. Soil mineral N was also measured on a composite sample from each treatment (2 cores per replicate). Sampling depths were 0 30 cm and cm, which reflected the maximum rooting depth of many vegetable crops. Other physical and biological indicators were also measured, including aggregate stability and hot water extractable carbon (HWEC); these analyses were only performed on samples collected from a single replicate (0 15 cm). Findings Table 6: What was the effect of cover crop selection on crop biomass, weeds and N dynamics immediately prior to incorporation? Cover crop 1 Crop biomass (t DM/ha) Proportion of weeds (%) 2 Crop N uptake (kg N/ha) 3 Soil mineral N (kg N/ha, 0 60 cm) N in the system (kg N/ha) 4 Chemical fallow Unsprayed fallow Ryegrass 7.6 < Oats 11.1 < Triticale 8.7 < Lupin Chicory Nemat ISCI Triticale + peas Vetch Wheat Vetch + oats Clover No measurements were taken in Phacelia and Buckwheat plots due to insufficient/no biomass. 2 Expressed as a percentage of the total dry biomass recorded. 3 Above-ground component only. 4 Calculated as the balance of N in the crop and soil (0 60 cm) at incorporation. The New Zealand Institute for Plant & Food Research Limited (2011) Page 9

16 General observations: Significant differences (P < 0.001, LSD = 4.1) in crop biomass were observed. The best performing options were generally cereals (oats and triticale in particular), mustards and ryegrass. These crops all established quickly in autumn, and also kept weed populations low by directly smothering weeds and reducing the availability of light at ground level. Control of weeds is an added benefit that can help growers to reduce herbicide use over time. Lower biomass cover crops were generally slower to establish. This resulted in much higher competition between the planted crop and weeds, which reduced overall productivity. Establishment of Phacelia was very patchy at the site, which meant it was not possible to quantify yields. Buckwheat established well but was killed following a severe frost in June. Performance of cover crops can vary widely depending on planting date, regional conditions and paddock suitability. Growers should therefore assess the relative performance and benefits of promising cover crops in their own fields to determine what works best for them. Cover crops can reduce leaching of mobile nutrients like N. In this trial there was a wide range in the amount of N in the soil and crop at incorporation, reflecting a combination of factors such as plot-to-plot variability, differences in the rooting depth of the cover crops (some may recovered N from below 60 cm), or possibly even legume activity (albeit conditions were quite cold for fixation). Readily available N (i.e. the sum of crop N and soil mineral N) in the chemical fallow plots at incorporation was on average 98 kg N/ha lower than the other cover crop options. The replacement cost of this N using urea fertiliser would be approximately $150/ha (at $720/t of urea). In most cases this represented at least half and in some cases all of the initial seed cost associated with the cover crop. Allowing weeds to grow (i.e. the unsprayed fallow) reduced N losses compared to the chemical fallow, but may increase weed pressure in subsequent crops due to reseeding. The New Zealand Institute for Plant & Food Research Limited (2011) Page 10

17 Table 7: What was the effect of cover crop selection on soil physical and biological indicators in the top 15 cm immediately prior to incorporation? Cover crop Aggregate stability (MWD, mm) 1 % of aggregates Hot water extractable > 1 mm 2 carbon (µg C/g soil) Chemical fallow Unsprayed fallow Ryegrass Oats Triticale Lupin Chicory Nemat ISCI Triticale + peas Vetch Wheat Vetch + oats Clover Represents the mean aggregates size after a standardised wet sieving test (an indicator of stability); MWD = mean weight diameter. 2 Represents the percent of aggregates that have a MWD of greater than 1 mm after the wet sieving test. General observations: Soil physical and biological indicators were highly variable. However, overall there were some trends that appeared to indicate a short-term benefit of many cover crops compared with the chemical fallow. Aggregate stability following wet sieving was lowest (poorest) in chemical fallow plots, with few big differences between the various cover crop options. In almost all cases MWD values were less than 1.5 mm (a threshold related to growth limitations). The percent of aggregates bigger than 1 mm following wet sieving was consistently higher where cover crops had been grown. Bigger aggregates in this context appeared to indicate improved soil aggregation and resilience to physical damage. Averaged across cover crop options, the percent of aggregates bigger than 1 mm was 42% compared with only 24% in the chemical fallow plots. The specific role of root structure was not assessed in detail in this work. Cover crops that have a taproot (e.g. mustards, chicory) may have less effect on attributes like aggregate structure compared to fibrous roots (e.g. cereals and grasses) but may help break soil compaction. This reinforces the need to select a cover crop that addresses the issue in each specific paddock. Hot water extractable carbon is an indicator of biological activity in the soil (representing the pool of soil C generally associated with microbes). Results were highly variable across cover crop options, but were generally higher than the chemical fallow. A key to the long-term success of cover crops that was not assessed in this or other trials was the retention of conditioning benefits during incorporation; unless cover crops are sprayed out early or there is a period during which the residue can decompose, The New Zealand Institute for Plant & Food Research Limited (2011) Page 11

18 some degree of cultivation is generally necessary. Cultivation has the potential to quickly undo the benefits associated with cover crops. Mulching the crop can assist incorporation and enhance the breakdown of residues. The specific role of cultivation in retaining soil physical structure is being addressed in a separate SFF project (Grass2Crop, SFF 08/038). Figure 6: There are clear differences in the rooting pattern of a cover crop like ryegrass (left, fibrous) and mustard (right, taproot). Other related case studies: A number of other demonstrations were supported during the course of this project, including cover crop comparisons in Hawke s Bay (four sites), Pukekohe (one site), Hamilton (one site) and Horowhenua (two sites). These were not reported in this summary given the nature of the trials (i.e. they were largely in support of a grower s own activity with cover crops). Key messages when selecting cover crops: There are a wide variety of cover crops available, many of which offer different benefits. Identifying the intended use is therefore a critical step in selecting the option that best addresses the issues in a specific paddock, thereby ensuring cover crop use is profitable longterm. To get the most from cover crops they should be fully integrated into the rotation (i.e. not an afterthought when a field is not being cropped) and their use sustained over the long-term (i.e. not once in a while). The frequency and duration of cover crops clearly requires a balancing act of the associated costs of resting land against the longer term provision of soil services. Many cover crops can be used to recycle nutrients from depth. Once these residues are incorporated, nutrients will be gradually released to the next crop. This improves nutrient use efficiency, and limits the risk of nutrient loss to the wider environment. To maximise this potential recycling effect, fertiliser inputs to cover crops should be minimised. Some cover is better than no cover as nutrient losses and soil physical damage can occur quickly. Similarly, soil condition can improve quickly due to cover crops, especially if incorporation and traffic practices limit any unnecessary cultivation and compaction. The New Zealand Institute for Plant & Food Research Limited (2011) Page 12

19 The cost of cover crops is often partially or fully offset by direct and measurable benefits such as the return of deep nutrients and reduced herbicide use. Other indirect benefits such as improved soil physical quality may be harder to value, but is clearly linked to the long-term productivity and resilience of the soil. Simple methods like a visual soil assessment should be used to measure and record changes in soil quality over time. This approach highlights the longer term impact of cover cropping practices and allows growers to consider changes to other management approaches (e.g. fertiliser rates, herbicide frequency). 3.2 Managing crop residues Cereals are an important break crop in a number of vegetable rotations. After harvesting the grain, growers can either bale the straw residue and sell it, or incorporate it to increase soil organic matter content. Weighing up the relative benefits of each approach can be difficult, especially as short-term profitability can often compete with longer term drivers of soil sustainability. Case study 3: Management of cereal straw Approach In 2010 a demonstration trial was set up to explore the short-term implications resulting from retaining cereal straw residue compared with the standard practice of baling and selling it. The site selected was in Pukekohe on a clay textured soil. Wheat was harvested for grain in February Shortly thereafter the two residue treatments were imposed; either the straw was cut close to ground level and baled (the standard approach in most of the field) or the straw was retained (a single strip in the middle of the field, 14 m wide x 60 m long). The difference in straw residue biomass was then recorded for each approach, and the NPK content of a representative subsample measured (nutrients are lost when the residue is baled and used elsewhere). The field was then conventionally cultivated ahead of onions, which were subsequently planted in June. All areas received the same crop management during the season. In January when the onion crop reached commercial maturity a yield determination was made from 12 plots taken from each prior residue treatment. The harvested area in each plot was 1.8 m 2, from which the count and mass of bulbs sorted according to key market sizes (bulb diameters of < 40 mm, mm, or > 70 mm) was recorded. Soil samples (0 10 cm) were also collected from these plots, and the organic profile (organic matter, available N, C:N ratio) and key aggregate size distributions following dry sieving measured. The New Zealand Institute for Plant & Food Research Limited (2011) Page 13

20 Figure 7: A comparison between the amounts of straw removed (left) and retained (right) after the grain harvest. Findings Table 8: What was the effect of residue management on the amount of biomass and nutrient exported? Indicator Straw removed Straw retained Residue returned to soil (t DM/ha) Residue exported (t DM/ha) Value from selling the straw (net $) 1 $615 - NPK exported in straw (kg/ha) 2,3 N 25 ($39) - P 05 ($21) - K 48 ($83) - Total value of lost nutrients ($/ha) $143-1 Based on a net return of $30 per bale and 200 kg straw/bale. 2 The NPK concentration of the straw residue was 0.62% N, 0.14% P and 1.19% K; other nutrients levels were 0.14% S, 0.09% Mg, 0.34% Ca and 0.02% Na. 3 Values in parenthesis reflect the replacement cost of each nutrient based on a cheap form of N, P or K fertiliser (urea, $720 per t; superphosphate, $345 per t; potassium chloride, $860 per t). These calculations assume full utilisation or availability of applied nutrients over time (in practice efficiency is likely to be much less). General observations: Retaining straw residue returned about 5.5 t DM/ha which would eventually be cycled into organic matter in the soil. Removing straw residue exported NPK and other nutrients from the field. The cost to replace lost NPK (about $143/ha) was much lower than the net profit from selling the straw ($615/ha). However, it difficult to place a financial value on the return of carbon to the soil in the longer term. Table 9: What was the effect of residue management on onion yield and size profile in the next rotation? Indicator Straw removed Straw retained Statistical significance 1 Total yield (t FW/ha) ns % target size range (40 70 mm) ns Average bulb mass (g/bulb) Bulb number (000/ha) ns 1 Relevant data were analysed using a paired t-test (n=24); ns = not significantly different at P < The New Zealand Institute for Plant & Food Research Limited (2011) Page 14

21 General observations: There was no significant effect of prior residue treatment on onion yields or the percent of yield that was within the target size category. Overall, onion yields were poor due to the very dry summer conditions. If the paddock had been irrigated the residue may have improved infiltration rates due to greater porosity. Incorporating the retained residue without increasing the number of cultivation passes posed some challenges. As a result, sowing the onion crop in these areas proved more difficult. Although not statistically significant, this effect appeared to be reflected by the difference in bulb number between plots at harvest (about 7% lower in the straw retained treatment). The smaller number of bulbs in this treatment was, however, countered by a trend towards bigger average bulb masses (i.e. the two components cancelled each other out to have no effect on yield). Mulching the straw first may be necessary to help ensure it is easily incorporated. Table 10: What was the effect of residue management on soil organic indicators and soil aggregate size in the next rotation? Indicator Straw removed Straw retained Statistical significance 1 Organic matter (%OM) Available N (kg/ha) ns C:N ratio ns Aggregate size distribution (%) 2 < 2 mm ns 8 2 mm ns > 8 mm 5 4 ns 1 Relevant data were analysed using a paired t-test approach (n=24); ns = not significantly different at P < Represents the percent of aggregates that have diameters within each noted size range following the standardised dry sieving test. General observations: Soil organic matter was significantly higher under the straw retained treatment (albeit the difference was very small). The remaining organic profile indicators were not significantly affected by the residue treatments. Overall, organic matter and available N were low, consistent with the long-term cropping history at the site. Organic matter content above 7% is generally referred to as optimum by the commercial testing labs. Aggregate size distribution following dry sieving was not significantly different between treatments. Most aggregates were small to very small, owing both to a history of regular cultivation and low organic matter content. The fact that there were no major changes in soil characteristics was not surprising given the type of residue being incorporated (i.e. heavily lignified, dry residue with a high C:N ratio). The benefits from such residues are more likely to be observed in the longer term. Potential winter benefits from retaining the straw residue were also not considered in this trial. These may include greater protection from rain and wind erosion, and better conditions for water infiltration. The New Zealand Institute for Plant & Food Research Limited (2011) Page 15

22 Key messages on managing crop residues: In many cases exporting straw residues is more profitable than retaining the residue. However, if there is no plan to return carbon back to the soil over the long term (either as crop residues or other break and green crops) soil organic matter content will typically decrease as a result of cultivation. Soils that have low organic matter contents can require more inputs to maintain productivity. A tool designed to help growers assess the direct economic cost of straw losses is available at: Using cover crops to recycle nutrients Adequate nutrient availability is an essential component of profitable cropping systems. Whether nutrients are released from reserves in the soil or applied as fertiliser, losses can be an ongoing concern for growers. For example, mobile nutrients like N can be leached following rainfall and irrigation, while less mobile nutrients like P and K can be buried during cultivation. Practices that recycle nutrients back to the soil surface can therefore offer significant gains in nutrient use efficiency while limiting potential losses to the environment. Earlier trials have shown that break crops like maize can be used to recycle nutrients from depth while improving soil physical characteristics. Maize is an ideal crop for this purpose because it has an extensive, deep root system (roots can penetrate to 1.5 m and in some cases deeper). Maize is also often recommended during pasture renewal phases in livestock systems because of its ability to help alleviate compaction damage and increase natural drainage. Case study 4: Maize to recover deep nutrients Approach One demonstration trial was conducted in Horowhenua during the summer season to quantify the amount of N that could be recycled from depth from a paddock with a long history of intensive vegetable production. Base fertiliser was broadcast to the field at a rate of 357 kg CAN/ha even though the soil had high background N concentrations. Maize ( Forage King ) was subsequently sown in late November at a population of 110,000 seeds/ha. The cost of the seed was approximately $110/ha (seed was untreated). One week after sowing, soil mineral N samples were collected from four permanent plots within the paddock. Each sample comprised four cores that had been composited for each depth increment (0 30 cm, cm, cm and cm). The crop received no additional inputs during the season. Immediately prior to the commercial harvest in mid-march crop biomass was measured in the four plots, and a subsample retained from each to determine DM content and total N concentration. Soil mineral N concentrations were also reassessed in these plots as per the same method at planting. Soil physical characteristics were not followed in this trial because there was no comparison to test against. The crop was then cut using a commercial forage harvester; instead of being used to make silage, maize was broadcast across the soil surface to recycle nutrients for the next vegetable crop. The New Zealand Institute for Plant & Food Research Limited (2011) Page 16

23 Figure 8: Finely-chopped maize residue was spread across the soil surface (left) by the forage harvester and later worked back into the soil by rotary hoe (right). This completed the cycling of nutrients from depth back to the soil surface where they could be more readily used by the next vegetable crop. Findings Table 11: What was the effect of maize on soil mineral N levels at planting and again at harvest? Time of sampling Soil mineral N (kg N/ha) by sampling depth 0 30 cm cm cm cm Total amount of N in the profile At planting 222 (96) At harvest % change 82% 32% 37% 33% 1 Includes the addition of 357 kg CAN/ha (96 kg N/ha) immediately prior to planting (it was assumed that fertiliser N did not leach prior to the initial sampling). General observations: Soil mineral N concentrations at planting were very high, reflecting both the current and historical application of N fertilisers in this field. In total, there was 385 kg N/ha in the top 120 cm of soil at planting, most of which was present as nitrate N (especially close to the soil surface). Additional N would have also been mineralised during the season, increasing the supply of plant-available N forms on top of what was initially measured at planting. The typical N demand of a crop like maize is much less ( kg N/ha). Of particular interest was the amount of mineral N measured between 60 and 120 cm at planting (132 kg N/ha). This N was well below the effective rooting depth of many vegetable crops. The replacement cost of N from this zone was equivalent to about $200/ha (at $720 per t as urea), higher than the initial cost of the maize seed. If N in the cm zone was also included (some crops struggle to get to this depth, especially when there are other soil physical constraints like compaction) the replacement cost was even higher ($250/ha). At harvest there was a clear effect of the maize crop on soil mineral N content. The crop itself removed approximately 350 kg N/ha, an amount that was much higher than expected. This was related to the very high N concentrations in the tissue (2% total N), which itself appeared linked to excess availability of N in the soil. Typical N concentrations for maize silage at harvest are 1 1.2% total N. The New Zealand Institute for Plant & Food Research Limited (2011) Page 17

24 The biggest change in soil mineral N content was in the top 30 cm, which decreased by 82% between planting and harvest. This was strongly influenced by the fact that the grower applied fertiliser even though soil N content was already high. As a result, plants would have used N preferentially from the surface layer as opposed to foraging as aggressively at depth. The percent change in the lower soil depths was therefore considerably less (32 37% lower at harvest than at planting). Key messages on using break crops to recycle nutrients: Maize and other deep-rooted break crops recycle nutrients like N from depths that are not typically explored by many vegetable crops. This provides an opportunity to increase nutrient use efficiency and reduce fertiliser costs. It also minimises the potential loss of mobile nutrients to the environment. A key factor in the ability of break crops to reduce soil N content at depth is the availability of N close to the soil surface. In most cases crops will preferentially take N from the surface layers because it is easier to do so. Eliminating or significantly reducing fertiliser applications during break crops is the best approach to maximise this recycling effect. High quality seed and/or high seeding rates are not always necessary for this purpose. The goal is to ensure there is a strong sink for nutrients, rather than a well-spaced, uniform crop. Lower grade seed mixes like that used in this demonstration trial provide growers with a cost effective option. If soil N content is still high at harvest the above-ground crop residue can be exported to other paddocks. The New Zealand Institute for Plant & Food Research Limited (2011) Page 18

25 4 Surface runoff Surface runoff is an ongoing concern for vegetable growers on flat and sloping land. Runoff occurs when water application (from rainfall or irrigation) exceeds the rate of water infiltration into the soil. Infiltration rate is influenced by intrinsic soil characteristics and managementrelated factors. For example, clay dominant soils typically have much slower infiltration rates than silt and sand dominant soils. Regular and intensive cultivation of many soils can also create impermeable pans or cause a general loss of soil physical structure that can further slow water infiltration. There are a number significant issues associated with surface runoff. Within a field, localised ponding can suppress growth and reduce crop yields. Ponded areas are also prone to disease outbreaks and compaction damage during routine crop management activities, both of which have the potential to limit grower profitability. Runoff can also carry fine sediment and mobile nutrients, which can lead to lost fertility and wider environmental impacts. Nutrients like N are also more easily leached and denitrified under saturated conditions. Figure 9: Poor infiltration can create surface ponding and runoff, both of which can have a significant effect on crop productivity (left). Ponded areas are also more prone to compaction damage by vehicle traffic (right). Despite the potential impacts, addressing surface runoff and ponding issues in the field is often overlooked because the associated costs are not well quantified. Earlier trials in the HIT project highlighted just how damaging this issue can be. In brief, ponding even for a short period caused major yield loss in onions, and was related to poor soil physical condition in affected areas. Similar observations have also been made for a range of other crops. The key emphasis has since shifted to demonstrating in-field mitigation options. In the past, sediment traps have been used in some regions to contain runoff and reduce offsite impacts. However, these traps require regular maintenance to be effective and do not resolve the issue of localised ponding in the field, or prevent surface runoff in the first instance. The purpose of new demonstrations was to test additional approaches that could be used to mitigate in-field ponding and runoff issues. Key areas of grower-interest centred on reducing the flow of water across the soil surface (i.e. giving more time for water infiltration) and either crop or chemical stabilisation of the soil (i.e. still allowing water to flow, but reducing the risk of soil and nutrients being lost in the process). It is important to note that other initiatives to improve soil physical condition can have a strong impact of infiltration rates; for example, building soil organic matter content through the use of cover crops can help with water infiltration, as can reducing the intensity of cultivation and extent of compaction. The New Zealand Institute for Plant & Food Research Limited (2011) Page 19

26 Figure 10: Sediment traps are used to limit offsite impacts (left) but require regular maintenance. The fine, poorly structured sediment collected from each trap must also be spread back to the paddock which can be costly (right). 4.1 Reducing the flow of water across the soil surface Uncontrolled movement of water down wheel tracks can result in significant soil erosion, either from the wheel track itself or from the edge of adjoining beds. The first step to reducing this risk is to identify and then control water that enters the paddock from surrounding areas (e.g. small streams, other surface drains or runoff from nearby paddocks). In many cases this can be achieved through the use of interception drains and bunds. Figure 11: Wheel track and bed erosion can be a significant issue in cultivated paddocks. In addition the movement of soil to the lowest area of the paddock (left), erosion can also have a significant effect on the productivity of the crop (right). However, there is still a risk of water flow within the paddock during rainfall and irrigation events. In these cases, some form of in-field control measures may be necessary to reduce water flow and limit soil erosion. A variety of practices have been advocated for the control of runoff in the field. The most well-known is wheel track ripping, which can increase the rate of water infiltration into the soil profile. However, some growers have reported that wheel track ripping can cause more soil erosion in certain situations, as it provides a preferential flow path for water down the wheel track and into the bed. Less invasive approaches such as wheel track diking may provide an alternative approach to control runoff. Diking involves creating small indents in the wheel track to increase water retention time and therefore infiltration, and has been used widely by growers in Australia and the US to increase water use efficiency during rainfall events. Early demonstrations in the HIT project have shown potential for wheel track diking to reduce surface runoff and sediment loss. For example, in a Hawke s Bay trial the volume of winter runoff from a trial bed that had been diked was reduced by over 90% compared with an undiked control. There were also clear effects on the amount of sediment that was lost in the runoff. The New Zealand Institute for Plant & Food Research Limited (2011) Page 20

27 Not diked Diked Figure 12: Comparison of control (left) and diked (centre) wheel tracks on the amount of surface runoff during a winter rainfall event. There was also a clear effect of wheel track practice on the amount of sediment in the runoff (right). Case study 5: Wheel track diking Approach A number of demonstration trials were established to highlight the opportunity to reduce surface runoff using wheel track diking. In most cases these trials focused on addressing the practical issues surrounding the use of this technology rather than quantifying the differences compared with standard wheel tracks. In total there were 13 demonstrations that covered a range of winter and summer cropping scenarios. Seven sites were located in the wider Pukekohe region, four in Horowhenua, and the remaining two in the Hawke s Bay. A variety of crops (onions, carrots, potatoes, spinach, cabbage and lettuce) and soil types (clay, silt and sand dominant) were covered. In each case the cooperating grower either used their own diking equipment. Findings General observations: Growers observed that wheel track diking reduced surface runoff and in-field ponding issues in most situations. Soil type itself did not appear to have a major effect on the practice, nor were there any crops in which wheel track diking was considered unsuitable by the grower. Figure 13: Wheel track dikes reduce water movement, preventing ponding in the lowest area of a paddock. The value of wheel track diking was demonstrated across a range of slopes. It is often incorrectly assumed that runoff is not an issue on flat land. However, any time water is applied at a rate faster than the soil can absorb it runoff is generated. In most cases this The New Zealand Institute for Plant & Food Research Limited (2011) Page 21

28 moves into small dips even if the paddock appears flat. Localised ponding even for short durations has a large impact on crop yield. Figure 14: Wheel track diking can be an effective on both flat and sloped paddocks. Growers found that a shallow ripper knife or shoe in front of the diking equipment helped to loosen the soil. This created more defined soil dikes, especially in wheel tracks that were heavily compacted. Figure 15: Loosening the soil ahead with a shallow knife or shoe improves the formation of soil dikes. Diking when the soil moisture content is high (i.e. close to or above field capacity) can cause significant damage to the soil and does not form complete dikes. As a general rule of thumb if the soil is too wet to undertake light cultivation practices then it is too wet to dike. Figure 16: If the soil is too wet for shallow cultivation it is too wet for wheel track diking. The New Zealand Institute for Plant & Food Research Limited (2011) Page 22

29 The number of times a crop should be diked during a cropping sequence will vary widely based on many factors. Dikes can break down over the winter months in particular, and may need to be reformed when conditions are favourable to maintain control. Stabilising the dikes with a secondary crop (e.g. grass, oats) can help, but introduces in an added complexity to the practice. Similarly, if every wheel track is used on a regular basis for crop management activities then dikes may need to be reformed. Diking spray rows is not generally recommended due to their regular use. Figure 17: Soil dikes may need to be reformed after heavy rainfall events or routine crop management activities. There are a wide variety of paddle configurations that can be used. Paddle configuration determines the height and distance between dikes. Agricultural engineers have experience in these issues, and can be consulted to optimise the performance of this tool. In many cases, diking equipment can be fitted to other implements, reducing the need for separate passes through the paddock. Figure 18: A number of design configurations are available to create soil dikes, most of which can be attached to other implements. The New Zealand Institute for Plant & Food Research Limited (2011) Page 23

30 Key messages on reducing the flow of water across soil surfaces: Reducing surface runoff and soil erosion from cultivated paddocks starts by preventing water from entering the paddock from adjacent areas. This can be achieved through the use of interception drains and bunds. However, control measures may also be necessary to prevent infield water movement. Wheel track diking is one approach that has been shown to be effective at reducing surface runoff and sediment loss across a wide range of conditions. Diking equipment is comparatively cheap, and can be attached to existing bed-forming implements. This reduces the need for further passes through the paddock to establish the dikes. The frequency of diking depends on the conditions in which it is used. In some cases one pass per season may be enough to minimise losses, whereas in others several passes may be necessary, between large rainfall or irrigation events. The equipment should not be used when the soil is too wet, as the dikes will be poorly formed and soil structure may be damaged. 4.2 Using secondary crops to reduce soil losses Planting secondary crops to stabilise the soil within an otherwise regular vegetable crop can reduce erosion. For example, in the Hawke s Bay some growers plant strips of maize within squash crops, while others in the region have used strips of cereals within onion crops to slow wind speeds and reduce wind erosion. An added benefit is that sensitive vegetable crops are protected from the abrasive effects of wind-blown soil. Figure 19: Wind erosion can be a major issue on some light or poorly structured soils following cultivation (left). One approach to limit these losses is to plant secondary crops in regular strips across the paddock to reduce the wind speeds (right). While these practices can provide clear benefits on soils that are prone to wind erosion they are less effective on soils prone to water erosion. In these situations, greater protection across the soil surface is generally necessary. An alternative approach then may be to plant a quickgrowing secondary crop of no economic value either in the wheel tracks (while the vegetable crop is growing) or across the whole bed (prior to the vegetable crop being planted). Cereals and grasses are ideal for this purpose, because they are cheap, quick to establish, produce a fibrous root system and are easily controlled by a variety of herbicides. Timing and ease of management are key factors that must be considered when using these approaches, as would any potential competitive effect on the more valuable economic crop. The New Zealand Institute for Plant & Food Research Limited (2011) Page 24

31 Case study 6: Crop stabilisers in the wheel track Approach In 2009 a demonstration trial was set up in Pukekohe to assess the potential of wheel trackplanted oats to reduce soil erosion in an otherwise standard potato crop. Soil texture was clay loam, and paddock slope was about 6. The site was prepared using conventional cultivation practices in the autumn, and subsequently planted with potatoes in early August. Immediately after the final pre-emergent herbicide application oats were sown by hand in two adjacent wheel tracks (i.e. one full bed) that were each 165 m long. The sowing rate was equivalent to approximately 30 kg seed/ha (the cost at this rate was approximately $50/ha). Adjacent to this test bed a comparable control plot was established that had no oats planted in either wheel track. At the bottom of the paddock silt monitoring traps were installed to collect any sediment lost from the two wheel tracks. The potato crop received standard fertiliser and herbicide management during the season, but was not irrigated. Figure 20: Wheel track-sown oats in early November as the potato crop was beginning to die back (left), and the sediment fences used to trap any eroded soil (right). Herbicide was sprayed over the entire paddock in early November, which reflected standard practice ahead of the potato harvest. To quantify the effect of wheel track practice on potato yields three small plot harvests were dug from each treatment. The harvested row length was 3 m, with each plot spaced approximately 40 m apart down the row. Plant count was recorded, as was the number and total fresh mass of tubers in each plot. The initial intent was to also measure the amount of sediment behind each silt trap. However, there were no major rainfall events during the trial that caused obvious movement of sediment. A soil sample was collected and analysed for basic fertility to assess any potential movement of soluble nutrients. It is important to note that sediment loss may have occurred between when the soil was first cultivated (i.e. autumn) and when the potato crop was subsequently planted (i.e. early August). During this period the field was fallow and exposed to the effects of winter rains. Findings Table 12: What was the effect of wheel track practice on the performance of the potato crop? Wheel track practice Number of plants (per ha) Crop yield (t FW/ha) Number of tubers (000/ha) Mean tuber mass (g/tuber) Fallow Oats Statistical 1 ns 0.05 ns ns significance 1 Relevant data were analysed using a paired t-test approach (n=6); ns = not significantly different at P < The New Zealand Institute for Plant & Food Research Limited (2011) Page 25

32 General observations: There was no effect of wheel track treatment on the number of potato plants at harvest. This was consistent with the fact that oats were sown after the potato crop had emerged. Tuber yields were significantly lower where oats had been grown in the wheel track during the season. Although neither component was significant, yield loss appeared to be primarily related to smaller tubers and to a lesser extent the number of tubers that set. This suggested there had been significant competition between the oats and potato crop for resources (e.g. water, nutrients, sunlight), even though they were grown in different zones (potatoes in the beds, oats in the wheel tracks). Competition between the two crops would make this approach unappealing to most growers. One option to reduce competition would be to spray out the wheel track crop earlier in the season using either a selective herbicide or through the use of a shielded sprayer. Table 13: What was the effect of wheel track practice on soil fertility in the silt traps? Wheel track practice Soil ph Olsen P (mg/l) MAF-K MAF-Ca MAF-Mg Organic matter (%) Fallow Oats General observations: Although the data were not replicated, there were no obvious differences in soil fertility between the two wheel track approaches. These findings were consistent with visual observations that indicated very little soil was lost during the monitoring period. Case study 7: Crop stabilisers across the paddock Approach In 2010 a demonstration trial was set up in Pukekohe to assess the opportunity to sow a temporary wheat crop to reduce sediment loss from a fallow paddock prior to planting. Soil texture was clay loam, and paddock slope was about 3.5. The paddock was cultivated in late April while ground conditions were still favourable, but was not scheduled to be planted in potatoes until July (the risk of soil loss during this period is often high). Wheat seed was broadcast in strips using a fertiliser spreader immediately prior to the final pass to form the potato beds in late April. There were two strips of wheat (each 24 rows wide) in the paddock, each with a matching control strip of the same width that was left fallow. Row length was approximately 320 m. Soil loss from the wheat and control strips was measured on several occasions during the winter using sediment tanks positioned at the bottom of the paddock. There were three tanks per strip, all placed in the wheel tracks. Because of the amount of sediment that moved, in most cases calculations had to be made using the depth and length of row that filled with soil rather than a straight measurement of soil mass inside the traps. These areas were cleared out by hand after each measurement date. In late June the paddock was sprayed out with herbicide, and potatoes were sown approximately 3 weeks later. The New Zealand Institute for Plant & Food Research Limited (2011) Page 26

33 Figure 21: A comparison of the amount of soil movement under the fallow (left) and wheat (right) treatments in early June. Findings Table 14: What was the effect of the wheat stabiliser crop on sediment loss during the winter? Winter practice 10 June Cumulative rainfall (mm) 1 Soil loss (kg DW/wheel track) 2 Fallow Wheat July Statistical significance 3 Fallow ns Wheat Represents the total amount of rainfall recorded between each monitoring date (19 May to 10 June, 11 June to 5 July). 2 The amount of soil measured was a relative indicator of the erosion loses under each treatment. 3 Relevant data were analysed using a paired t-test approach (n=6); ns = not significantly different at P < General observations Total rainfall measured at the closest NIWA weather station (Pukekohe, about 16 km from the trial site) was 354 mm during the study. The maximum daily rainfall during monitoring was 47 mm, which fell in late June. When the wheat crop was sprayed out there was approximately 1.3 t DM/ha in above-ground biomass and the standing height of the residue was about 15 cm. Broadcasting wheat across fallow soil reduced soil loss by approximately 38% between May and June, and by approximately 26% between June and July. Although neither finding was statistically significant, this appeared to reflect the small number of monitoring tanks and the high degree of spatial variability in runoff patterns. Visually, there was a clear benefit from the wheat treatment in most rows. The smallest losses were observed where wheat had successfully established in the wheel tracks as well as on top of the hills (as opposed to the hills only). Practices that ensure an even spread of seed during the bed forming process should increase the degree of control achieved. Wheat residue was quick to die back after herbicide application in late June. There was no delay or problems encountered during the subsequent planting of potatoes 3 weeks later. No final yield measurements were taken in the potato crop as there was no evidence to suggest there would have been any negative effect resulting from the small amount of wheat residue. ns The New Zealand Institute for Plant & Food Research Limited (2011) Page 27

34 Key messages on using secondary crops to reduce soil losses: Secondary crops can be used to stabilise soils and reduce erosion during periods of high risk. Planting these crops across an entire paddock appears to be an easier system to manage and more effective approach than only planting in the wheel tracks. Quick establishment of the secondary crop is a critical factor in reducing erosion losses. To minimise the period in which there is no cover, the secondary crop should be sown as close to the final cultivation pass as practical. A decision on the best time to terminate the secondary crop ahead of planting can be made depending on the conditions and amount of growth that has occurred. Because there is only a small amount of residue there is unlikely to be any delay planting the subsequent vegetable crop. The cost of this approach is low. In these demonstrations old seed that was left over from a cereal harvest was used. Additionally, the paddocks are typically sprayed with herbicide ahead of planting to control winter weeds, so there is no additional cost to control the secondary crop. Another potential benefit associated with the use of a secondary crop is the opportunity to return carbon back into the soil. Although the amount of biomass produced is small, there is no additional cultivation required to work it in. An increase in soil organic matter content over time can assist with the development of better soil structure. 4.3 Using chemicals to stabilise the soil Chemical stabilisers like polyacrylamide (PAM) are used widely in the roading construction industry to reduce sediment loss to waterways. The products work by flocculating soil particles together (clay and silt in particular) using differences in ionic charge, which causes them to rapidly settle out of suspension. Additional benefits from using these products have also been reported in the dairy and horticultural industries. For example, in the South Island farmers have added PAM products to irrigation water to improve infiltration under dairy pastures. In Australia, they have been used for similar purposes in grape vineyards and potato paddocks. Comparatively little is known about the performance of these products on cultivated cropping land in New Zealand. Case study 8: Chemical stabilisers across the paddock Approach Before testing PAM in a commercial setting two background studies were performed. An off-theshelf product was recommended by a New Zealand supplier for this purpose (Soilfix Magnafloc X135).The first study was to determine the likely concentration of PAM necessary to settle out soil in a water column. The soil used for this test was a clay loam, which was mixed at a high loading rate with water to simulate winter runoff from a cultivated field. Three concentrations of PAM were added to the soil-water suspension (5 ppm, 10 ppm and 30 ppm); a water only control was also included. Settling velocity was recorded over a 24-hour period after the suspensions were thoroughly mixed. All PAM treatments settled sediments out of solution much quicker than the control. Approximately 90% of suspended sediments had settled within the initial 30 seconds across all rates of PAM, whereas there were still suspended sediments after 24 hours in the water only control. The New Zealand Institute for Plant & Food Research Limited (2011) Page 28

35 Figure 22: Effect of the chemical stabiliser on the amount of sediment still in suspension after 30 seconds (left) or 24 hours (right). In each case, the cylinders from left to right are the water only control (no PAM), PAM at 5 ppm, PAM at 10 ppm and PAM at 20 ppm. A second study was then performed to assess any potential toxicity effect of the product on onion seedlings. Control (water only) and three PAM concentrations (5 ppm, 10 ppm and 20 ppm) were sprayed directly to trays of onion seedlings (six per tray) until there was runoff. Seedlings were then left outside to simulate a field environment and observed daily. After 1 week there were no visual signs of leaf damage on any of the treatments. The dry mass of plants at this time was 1.1 g, 2.1 g, 1.1 g and 0.8 g for the control, 5 ppm, 10 ppm and 20 ppm PAM rates respectively. Variation in biomass reflected the small number of test plants. The implications of these initial tests were that a low concentration of PAM (5 ppm) was effective at dispersing soil from water. There was also no evidence of a negative effect on onion seedlings. A scoping trial was subsequently established in Pukekohe to test the potential of PAM to reduce soil erosion from a cultivated paddock. Soil texture was clay loam, and the paddock slope was about 3. The site was prepared using conventional cultivation practices in the autumn, and planted with onions in late May. PAM was applied to the wheel tracks and bed tops of six onion beds in early June (before the crop emerged). Application was by battery powered knapsack at a rate of 5 kg product/ha (the product cost at this rate was approximately $68/ha). Two rows were then skipped, followed by six rows of control onion beds that received no PAM. Row length across the trial area was 100 m. Sediment monitoring tanks were immediately installed in the wheel tracks (five tanks per practice) at the bottom of the paddock. The trial concluded in mid-august, when sediment losses were assessed. Figure 23: Soil erosion from winter crop onion paddocks can be an issue due to the fine seed bed that is prepared (left). Applying a chemical stabiliser to the soil was seen as one approach to potentially reduce these loses (right). The New Zealand Institute for Plant & Food Research Limited (2011) Page 29

36 Findings General observations Total rainfall measured at the closest NIWA weather station (Pukekohe, about 2 km from the trial site) was 391 mm during the monitoring period. The maximum daily rainfall during monitoring was 47 mm, which fell in late June. Despite regular rainfall, there was no evidence of soil movement under either treatment. This likely reflected the intensity of rainfall (i.e. no sustained or major rainfall events) and the fact that some sediment loss may have already occurred prior to the treatments being applied. Slope may also have been an issue; however, the grower indicated that the paddock had been prone to soil erosion losses in the past. Further studies are therefore necessary to quantify the effect of PAM on erosion losses from cultivated soils. An additional benefit that was not assessed in the current case study was the potential of the product to increase water infiltration rates. This may be beneficial during summer irrigation practices where surface runoff can be an issue. It is important to note that the application of PAM on such a small scale was challenging due to its viscosity. The use of commercial sprayers would allow for a much more uniform application pattern in future studies. The rate of degradation under UV light also needs to be assessed, as this will influence the length of protection possible in exposed field conditions. Key messages on using chemical stabilisers to reduce soil loss: PAM showed potential to settle out sediments in laboratory conditions and appeared to have no negative effect on a small test crop sample. However, any potential field benefits were unclear because there was no sediment movement at the trial site. Further field testing is necessary. Other opportunities to use PAM in vegetable production could also be explored. For example, instead of spraying the product across the whole field its use could be targeted in sediment traps only. These areas collect and hold runoff for extended periods to allow fine sediments to settle out. By increasing the rate at which sediment settles, the number and frequency of traps required in each paddock could potentially be reduced, which may increase the effective cropping area of a paddock. There are also a variety of other products (chemical and non-chemical) that are being used by erosion control specialists in related industries. With some adaptation, some of these may provide dual benefits for vegetable growers. The New Zealand Institute for Plant & Food Research Limited (2011) Page 30

37 5 Cultivation and soil compaction There are many reasons why growers cultivate the soil ahead of planting vegetable crops. These reasons can include: to manage crop residues, control weeds, prepare seed beds, aerate the soil, incorporate fertiliser, or alleviate compaction. However, there are a number of issues that can arise from regular and intensive cultivation, which can reduce the profitability of cropping systems. Regular and intensive cultivation affects soil physical quality by repeatedly breaking down soil aggregates. In the process organic matter is also exposed to decomposition, reducing the ability of these aggregates to reform naturally (organic matter supplies humus, a binding agent in the soil). Poorly structured soils are generally less productive and more prone to erosion losses. Figure 24: Crop productivity in compacted soils is often poorer, as roots are unable to fully explore the profile to extract nutrients and water (left). By contrasts, where this constraint does not exists roots have much bigger resource to utilise (right). The amount and type of cultivation and the conditions under which it is undertaken are key drivers in minimising the short- and long-term impacts of cropping on the physical quality of soil. However, adjusting these practices is not always an easy task, as cultivation approaches are often closely linked to the amount of compaction in a paddock. For example, where heavy compaction has been caused by vehicle traffic or subsurface pans have been formed by tillage implements, some form of remedial cultivation is often necessary. In many cases then, combined approaches are necessary to firstly minimise the amount of compaction so that the amount and intensity of cultivation can be adjusted. The New Zealand Institute for Plant & Food Research Limited (2011) Page 31

38 Figure 25: A variety of cultivation practices are used by growers. In many situations the amount and type of cultivation used is directly proportional to the extent of compaction and the prior condition of the soil. 5.1 Controlled traffic farming approaches Controlled traffic farming (CTF) methods have been used widely overseas to increase the efficiency and profitability of cropping operations as well as to improve soil physical quality. The guiding principle of CTF is to keep all field traffic on permanent wheel tracks, thereby reducing compaction damage to the areas that are used to grow crops (i.e. beds). In addition to a wide range of energy use benefits, controlling traffic can have a direct impact on the productivity of the crop by improving soil structure. It can also reduce the need for intensive remedial cultivation at the end of the season to address the impacts of random traffic (particularly problematic at harvest). Case study 9: Controlled traffic farming Approach In 2009 a demonstration trial was established in the Pukekohe region to assess the potential of CTF approaches within an intensive vegetable rotation. The work was a collaborative effort with the SFF-funded Advanced Farming Systems project (SFF 08/111) which focused on efficiency gains (fuel use, fleet management, labour) and soil carbon storage resulting from CTF. The paddock selected had been cropped regularly for a number of years according to a standard 3- year rotation that included onions, potatoes and a cereal break crop. Soil texture in the paddock was clay loam, and the slope was approximately 5.5. At the start of the trial the paddock was split in two; half (3.5 ha) was to be managed according to the conventional cultivation and traffic approach of the collaborating grower, and the other half (3.5 ha) according to CTF principles. The trial commenced with onions that were planted in June 2009 and harvested in mid-january The biggest differences during this first crop sequence related to the initial formation of the beds and how the crop was harvested. Potatoes were subsequently planted in June 2010 and harvested in late October. During this second crop sequence, there were major differences in the amount of cultivation necessary to prepare beds for planting. This was due to the compaction effect caused by random traffic in the conventional area when onions were harvested. Harvest practices were also different under each approach. A summary of the key differences between the two systems over the course of the trial is provided below. Both the onion and potato crops were otherwise managed according to standard practice for the area. The New Zealand Institute for Plant & Food Research Limited (2011) Page 32

39 Table 15: What was the difference between conventional and CTF approaches during the trial? Activity Period Conventional CTF First crop sequence - onions Primary cultivation February March Disk, subsoil rip, tyne plough, power harrow No difference Bed formation April Mechanical bedder, roller Mould, mechanical bedder Crop management June January As per standard practice Harvest management January Random traffic, including vehicles on bed tops Second crop sequence potatoes Primary cultivation February March Disk, subsoil, rotary hoe No difference Controlled traffic, all vehicles confined to wheel tracks only Subsoil (beds only) and wheel track renovation Bed formation April Struck/ridging No difference Crop management June October As per standard practice Harvest management October Random traffic, including vehicles on bed tops No difference Controlled traffic, all vehicles confined to wheel tracks only In each sequence a variety of crop and soil measurements were collected to quantify the impact of the management changes. In all cases these measurements were confined to the same zone in the paddock to reduce potential sampling variability; a zone was 15.5 m wide and comprised either 9 beds (onions) or 18 beds (potato). There was one spray track in the middle of each zone. In the first sequence, onion yields were measured immediately prior to the commercial harvest (i.e. before differences in harvest traffic approaches had been implemented). These were measured in nine plots within each zone (each comprising three separate 2 m row sections). The total number of bulbs and their fresh mass was recorded and average bulb mass calculated. Soil physical characteristics (bulk density, mean aggregate size following dry sieving, and indicators of aggregate stability following wet sieving) were also assessed in each plot. The sampling depth was 0 15 cm, and there were three cores (for bulk density) or four spade x spade samples (for aggregate size and stability) collected per plot. Mean aggregate size and the associated aggregate distribution within key size ranges (> 9.5 mm, mm, < 1mm) was reassessed after the onions had been commercially harvested and the beds prepared for the next crop (i.e. after the differences in harvest traffic approach and remedial cultivation had been implemented). The same sampling depth was retained, and there were four spade x spade samples collected per plot. During the second sequence the effect of each management approach on soil erosion over winter was assessed. For this measurement, six sediment monitoring tanks were installed within each management zone. The tanks were positioned in the wheel tracks at the bottom of the paddock, and were spaced evenly around the spray track (using the same pattern in both zones). Spray tracks were not followed because of the regular traffic in these areas. Monitoring began in mid-may, and concluded in September. Potato yields were measured immediately prior to the commercial harvest in nine plots collected from each management approach (each comprising two separate 6 m row sections). The The New Zealand Institute for Plant & Food Research Limited (2011) Page 33

40 number of plants in the harvested area was recorded, and then a small plot lifter was used to dig the tubers. Tuber count and fresh field mass was recorded. Potatoes were then cleaned thoroughly with water to simulate the commercial packing process, and the fresh clean mass recorded. Following the commercial harvest in the paddock soil physical characteristics were assessed under each management approach. Penetration resistance was recorded as the average of 40 readings made using a hand-held penetrometer at 0 15 cm and cm sampling depths. Aggregate size distribution was also assessed in each plot. Sampling depth was 0 15 cm and there were three spade x spade samples collected in each plot. Findings Table 16: What was the effect of conventional and CTF approaches on the productivity of onions? Management approach Total yield (t FW/ha) Number of bulbs (000/ha) Mean bulb mass (g/bulb) Conventional CTF Statistical 1 ns significance 1 Relevant data were analysed using a paired t-test approach (n=18); ns = not significantly different at P < General observations: Management approach had no significant effect on onion yields. These observations were similar to those reported by the grower who also monitored yields in the wider paddock using pack-out data (equivalent to about 51 t/ha for both practices). There were significant but contrasting effects of management approach on bulb number and bulb mass. In the conventional zone there were a greater number of smaller bulbs, whereas under CTF there were fewer but bigger bulbs. These differences appeared related to the initial sowing populations achieved in each area (slightly lower in CTF due to the bed forming equipment that was used). Table 17: What was the effect of conventional and CTF approaches on soil physical indicators in the top 15 cm immediately prior to the commercial harvest? Management approach Soil bulk density (g/cm 3 ) Mean aggregate size (MWD, mm) 1 Aggregate stability (MWD, mm) 2 % of aggregates > 1 mm 3 Conventional CTF Statistical 4 ns ns ns ns significance 1 Represents the mean aggregate size after a standardised dry sieving test; MWD = mean weight diameter. 2 Represents the mean aggregates size after a standardised wet sieving test (an indicator of stability). 3 Represents the percent of aggregates that have a MWD of greater than 1 mm after the wet sieving test. 4 Relevant data were analysed using a paired t-test approach (n=18); ns = not significantly different at P < General observations: There was no notable effect of management approach on soil bulk density, aggregate size or aggregate stability immediately prior to the commercial harvest of onions. At the The New Zealand Institute for Plant & Food Research Limited (2011) Page 34

41 time of measurement the only difference between the two approaches had been the bed forming practice. Figure 26: Under conventional management the wheel tracks and bed tops were used to harvest onions (left), whereas under CTF management only the wheel tracks were used (right). Table 18: What was the effect of conventional and CTF approaches on soil physical indicators in the top 15 cm after the harvest and remedial cultivation? Management approach Mean aggregate size (MWD, mm) 1 Aggregate size distribution 2 > 9.5 mm (%) 9.5 1mm (%) < 1mm (%) Conventional CTF Statistical ns significance 1 Represents the mean aggregate size after a standardised dry sieving test; MWD = mean weight diameter. 2 Represents the percent of aggregates that have diameters within each noted size range following the standardised dry sieving test. 3 Relevant data were analysed using a paired t-test approach (n=18); ns = not significantly different at P < General observations: Mean aggregate size was much smaller under CTF, primarily due to fewer big aggregates (> 9.5 mm) and more small aggregates (< 1mm). This appeared to highlight the prior effect that harvest traffic (i.e. random or controlled) and/or the resulting cultivation practices (i.e. full cultivation or minimal cultivation) had on soil physical characteristics. Interestingly, CTF resulted in a smaller aggregate size despite considerably less cultivation. Figure 27: Mean aggregate size following the onion harvest and subsequent cultivation was bigger under conventional management (left) compared to CTF management (right). This appeared to reflect the impact of wheel traffic at harvest in particular. The New Zealand Institute for Plant & Food Research Limited (2011) Page 35

42 Table 19: What was the effect of conventional and CTF approaches on soil erosion during winter? Management approach 9 June Cumulative rainfall (mm) 1 Soil loss (kg DW/wheel track) 2 Conventional CTF 56 5 July Conventional CTF 15 8 September Statistical significance 3 Conventional ns CTF 7 1 Represents the total amount of rainfall recorded between each monitoring date (19 May to 9 June, 10 June to 5 July, 6 July to 8 September). 2 The amount of soil measured was a relative indicator of the erosion loses under each treatment. 3 Relevant data were analysed using a paired t-test approach (n=12); ns = not significantly different at P < Zero values for all conventional tanks on this date prevented the use of a paired t-test approach. - 4 ns General observations: Total rainfall measured at the closest NIWA weather station (Pukekohe, about 17 km from the trial site) was 681 mm during the monitoring period. The maximum daily rainfall during monitoring was 47 mm, which fell in late June. Erosion loses were particularly variable from tank to tank. However, overall there appeared to be a consistent trend towards higher soil loses under CTF (albeit not significant). This could have been related to the differences in aggregate size following the previous onion crop sequence (where CTF had a higher proportion of smaller-sized aggregates). Small aggregates are more prone to particle detachment during heavy or sustained rainfall events. It is important to note that because less cultivation is required under CTF approaches, soil organic matter should gradually increase. In turn, this should improve the natural aggregation of small aggregates, making them more resilient to loss. Figure 28: There appeared to be more sediment lost from the CTF approach (left) than the conventional approach (right) during the winter months, an observation possibly linked to smaller aggregate sizes under CTF. The New Zealand Institute for Plant & Food Research Limited (2011) Page 36

43 Table 20: What was the effect of conventional and CTF approaches on the productivity of potatoes? Management Plant Total yield (t FW/ha) % of soil Number of Mean approach population (per ha) In the field After cleaning attached after lifting tubers (per ha) tuber mass (g/tuber) Conventional CTF Statistical 1 ns ns significance 1 Relevant data were analysed using a paired t-test approach (n=12); ns = not significantly different at P < General observations: There were no differences in the number of potato plants that established under either management approach. However, total yields (in the field and after cleaning) were significantly lower under CTF. This was due to smaller tuber size rather than a decrease in the number of tubers that set. Differences in potato yields were strongly related to weed control issues at the site. The CTF area appeared to miss a key herbicide application during the season, which in turn had a major impact on crop productivity. This was unrelated to the treatments themselves. A key indicator of interest was the amount of soil still attached to the potatoes after lifting (better structured soils may separate more easily from tubers during harvest). There was clear evidence that CTF reduced the adhesion of soil to potatoes by more than 40% compared with the conventional approach. Less soil debris reduces the cost of transporting and washing the crop. In a separate assessment, pack out and quality indicators were measured by running commercial-scale batches of potatoes from the trial site through a local packing facility. These tests confirmed that pack out was much higher under CTF (approximately 91%) than under the conventional approach (approximately 78%). The primary difference between the two approaches was the incidence of powdery scab, which was almost twice as high in conventional batches. It was not clear whether this reflected a true difference between the management approaches or an existing pattern of disease pressure within the paddock. As was the case for the field assessment, similar reductions in the amount of soil still attached to potatoes from CTF were also observed in the packing facility. Figure 29: Conventional (left) and CTF (right) areas of the paddock immediately ahead of the potato harvest in late October. There was a clear difference in weed density between the two practices due to a missed spray in the CTF area. The New Zealand Institute for Plant & Food Research Limited (2011) Page 37

44 Figure 30: Soil that is still attached to potatoes after harvest represents added inefficiency within the system (left). Not only is it transported back to the packing shed, it must be removed (right) and then later disposed of sustainably. Under the CTF approach there was less soil still attached to the potatoes after lifting. Table 21: What was the effect of conventional and CTF approaches on the area of paddock that was subject to wheel damage during harvest and the effect this had on measures of soil compaction? Management approach Area affected by Penetration resistance (MPa) 2 wheel damage (%) cm cm Conventional CTF Based on two 20 m transects across in each management approach. At 20 cm increments a yes/no response was recorded to indicate whether that soil surface had been subject to wheel traffic. 2 Penetration resistance values were adjusted for soil moisture content. General observations Uncontrolled traffic approaches at harvest resulted in direct wheel damage to a very high proportion of the soil surface within conventional plots. By contrast, the degree of damage was considerably less under CTF where only the wheel tracks were used at harvest. Penetration resistance was measured as an indicator of the extent of soil compaction caused during the harvesting operations. These values only served as a relative indicator of management differences (due to recent soil disturbance); however, the readings still demonstrated that there was more compaction of beds under the conventional management approach than CTF, both in the surface and subsurface soil layers. Figure 31: Most of the soil surface was directly driven over by equipment in the conventional plots during harvest (left). By contrast, only the wheel tracks were driven over in the CTF plots at harvest (right). The New Zealand Institute for Plant & Food Research Limited (2011) Page 38

45 Table 22: What was the effect of conventional and CTF approaches on soil physical indicators in the top 15 cm after the harvest? Management approach Mean aggregate size (MWD, mm) 1 Aggregate size distribution 2 > 9.5 mm (%) mm (%) < 1 mm (%) Conventional CTF Statistical 3 <0.001 < <0.001 significance 1 Represents the mean aggregate size after a standardised dry sieving test; MWD = mean weight diameter. 2 Represents the percent of aggregates that have diameters within each noted size range following the standardised dry sieving test. 3 Relevant data were analysed using a paired t-test approach (n=18); ns = not significantly different at P < General observations: Mean aggregate size was significantly smaller under CTF, an observation that was due to fewer big aggregates (> 9.5 mm) and more medium to small aggregates (< 9.5 mm) than the conventional approach. Bigger aggregates in the conventional plots likely reflected the greater amount of traffic across these beds during the preceding harvest operations (i.e. aggregates were consolidated together due to pressure). It is important to distinguish the difference between larger aggregate size caused through these damaging processes to that achieved by reduced cultivation or increased organic matter retention (i.e. natural aggregation). Cultivated aggregates Cultivated aggregates Natural aggregates Compacted aggregates Figure 32: Large aggregates caused by compaction are of low value to crops compared to soil that has not been compacted (left). By contrast, large aggregates formed through natural aggregation provide a more productive soil base than that achieved through cultivation (right). Case study 10: Permanent beds Approach In October 2009 a demonstration trial was established in the Pukekohe region to assess the potential of permanent bed approaches to improve soil physical quality. This demonstration was also a collaborative trial with the Advanced Farming Systems project (SFF 08/111). The paddock selected was used for short term mesclun crops which have a typical duration of only 4 8 weeks (depending on the season and mesclun type). There are typically six or seven crop cycles per year, which results in a very high cultivation frequency. Soil texture in the trial paddock was clay loam. At the start of the trial the 1 ha paddock was split in two. Half of the paddock was to be managed according to the conventional cultivation approach of the collaborating grower, which The New Zealand Institute for Plant & Food Research Limited (2011) Page 39

46 included full cultivation by rotary hoe (several passes were often required to prepare a suitable seed bed) after every crop cycle and then reformation of the bed. In the other half of the paddock, beds were left intact after every crop and only reworked using a shallow rotary hoe and shaping tool. The established wheel tracks were therefore retained from crop to crop. Figure 33: Under the conventional approach beds were fully cultivated and then reformed between each crop cycle (left), whereas under the permanent approach beds were left intact but lightly cultivated between each crop cycle (right). Soil aggregate characteristics were assessed in April 2010, which followed approximately 6 months of the different cultivation approaches (three crop cycles). Eight samples (each a composite of two spade x spade samples spaced 10 m apart) were collected from each half of the paddock using a consistent sampling pattern across the paddock. Sampling depth was 0 10 cm. Mean aggregate size and the associated aggregate distribution within key size ranges (> 9.5 mm, mm, < 1 mm) was subsequently measured by dry sieving procedures. No further measurements were made at the demonstration site because the grower converted the entire paddock into a modified permanent bed approach. Findings Table 23: What was the effect of cultivation practice on soil physical structure in the top 10 cm after six months? Cultivation practice Mean aggregate size (MWD, mm) 1 Aggregate size distribution 2 > 9.5 mm (%) mm (%) < 1 mm (%) Conventional beds Permanent beds Statistical ns significance 1 Represents the mean aggregate size after a standardised dry sieving test; MWD = mean weight diameter. 2 Represents the percent of aggregates that have diameters within each noted size range following the standardised dry sieving test. 1 Relevant data were analysed using a paired t-test approach (n=16); ns = not significantly different at P < General observations: Mean aggregate size was larger where permanent beds were used, an observation that was due to more big aggregates (> 9.5 mm) and fewer medium to small aggregates (< 9.5 mm). This appeared to directly reflect the differences in cultivation practice between the two approaches (i.e. more intensive cultivation caused a greater loss of natural aggregation). The New Zealand Institute for Plant & Food Research Limited (2011) Page 40

47 Key messages on using CTF and permanent beds to reduce compaction damage and cultivation: CTF and permanent bed approaches provide growers with an opportunity to limit compaction damage and reduce the amount of cultivation necessary between crop sequences. This can have a major impact on fuel use efficiency, as well as on soil physical quality. Cultivating wheel tracks is an energy intensive process that results in large poorly structured aggregates being brought to the soil surface. Depending on the extent of compaction, these aggregates often require several passes with powered cultivation implements to break them down and prepare a suitable seed bed for the next crop. In many instances, this same degree of cultivation is in excess of what is actually required for soil that was previously within the bed top area. Over time this can result in a gradual loss of important soil physical qualities that are associated with infiltration, drainage and aeration. The suitability of CTF and permanent bed approaches will depend on the cropping system, and in particular, the variety of crop types that are grown and the type of equipment already in the fleet (or the ease with which modifications can be made). 5.2 Designated traffic zones to limit compaction damage CTF and permanent bed approaches allow growers to reduce widespread compaction damage and the need for remedial cultivation. However, one of the potential challenges associated with CTF is that for maximum benefit all crop sequences should use the same bed configuration (or multiples off); this ensures that the compacted wheel track in one crop remains the compacted wheel track in the next. In some systems this can be difficult to achieve due to the wide variety of crops grown and the existing equipment and implements in the fleet. Reconfiguring the planting arrangement can be a costly and therefore unappealing option. An alternative approach in this situation is to establish designated traffic lanes for the highest risk activities such as in-season crop management and harvesting operations. Figure 34: Uncontrolled traffic at harvest can cause major damage to soil physical structure (left and right). Soil aggregate condition in affected areas can remain poor even after remedial cultivation, which can impact productivity in the next crop. The New Zealand Institute for Plant & Food Research Limited (2011) Page 41

48 Case study 11: Designated traffic zones Approach Several sites were followed in Horowhenua where designated traffic lanes were implemented by the grower. These areas were established at the start of each crop sequence and were used for all in-season spray passes as well as the final harvest. This was possible because the crops were hand harvested and either packed directly into field crates or put onto conveyor belts to a central packing platform. No measurements of soil physical quality were taken at these sites; instead, the general observations were noted on the potential of the approach. Figure 35: Designated traffic lanes can be used to control traffic within the field, especially at high risk times like harvest. In some cases these area are not planted at all (left), whereas in others a lower planting density is used to reduce the loss of effective cropping area (right). Key messages on using designated traffic zones: Designated traffic lanes provide an effective means of confining compaction damage, especially if the crop is being harvested when the soil is wet. Because these areas are routinely used during the season they also tend to be more consolidated, which can actually reduce delays accessing fields following heavy rainfall events. These traffic areas can be retained from crop to crop, reducing the need to cultivate heavily compacted or damaged areas. This provides growers with an opportunity to better match the amount of cultivation to the soil conditions in the rest of the paddock. Planting configurations can be adjusted in the rest of the paddock to ensure that there is no overall loss of crop yields. The New Zealand Institute for Plant & Food Research Limited (2011) Page 42