Take home messages

• Applying urea onto saturated (but not flooded) soil before or during waterlogging may mitigate some of the effects of waterlogging on the growth and biomass of barley and canola however the effect on grain yield requires further analysis.
• Raised beds provide further increases in growth.
• Mineral nitrogen (N) deeper than 30cm should be considered inaccessible to crops that have experienced waterlogging, because waterlogging delays root extension.

Background

Waterlogging occurs in most years in the high rainfall zone (HRZ) and is a cause of reduced yield and a constraint to crop choice. In this paper we refer to waterlogging as both a flooded soil where water is visible above the surface, and soils where there is a water table as between the surface and 10cm deep. A study in commercial wheat paddocks in the HRZ in 2017 and 2018 found a yield reduction in wheat of up to 38% in poorly drained areas (Robertson et al. 2019). A follow-up study in 2021 where waterlogging was imposed by irrigation, found a yield reduction of 38% in wheat and 14% in faba beans. To fill gaps in the other major high rainfall crops of barley, canola and faba beans, in 2023 GRDC contracted the University of Tasmania and Agriculture Victoria Research to conduct a project ‘Understanding the impacts of waterlogging on barley, canola and faba bean’, with components of field experimentation, improvement of existing crop growth models, and use of the improved models in an economic assessment of solutions, such as drainage. This paper reports on field experimentation from UOT2306-001RTX and previous projects to assess the magnitude of loss and the potential of timely N application to mitigate such losses. It is based on maturity biomass data, as grain yield data are incomplete at the time of writing, but these will be presented at the Update.

Method

Study 1 – Timing and depth of waterlogging

To quantify the timing and depth of water tables in a typical paddock in the HRZ, water height recorders (Odyssey^® Dataflow, Christchurch, New Zealand) were installed in a continuously cropped paddock at Wickliffe between 2019 and 2021. Six positions were monitored in 2019, ranging from well-drained to poorly drained, ten positions in 2020 and twelve positions in 2021. The paddock was sown to wheat in 2019, canola in 2020 and wheat in 2021. This was part of the Victorian Grains Innovation Partnership project VGIP2A.

Study 2 – Options for spring sowing and late season waterlogging

To quantify the effects of waterlogging on the yields of canola, barley and faba beans, and the effects of recovery N applications, a field experiment was sown at Hamilton on 6 June 2023. However, 145mm of rain fell over the following 3 weeks, leading to seed burst and minimal plant establishment. The experiment was resown 4 September as soon as the paddock was trafficable, to compare options for resowing in spring with a full soil profile, while also quantifying the effects of imposed late season waterlogging on growth and yield. The soil was a Chromosol derived from basalt, with 80 kg/ha of mineral N in the 0–60cm zone and 134 kg N/ha in the 0–100cm zone. Waterlogging was imposed by drippers to two treatments between 19 October and 16 November. The drippers were operated on timers, a typical cycle being on for 40 minutes, four times per day. One of these treatments received 90 kg/ha of N as urea immediately before waterlogging (N before WL), while the other received it at the end of waterlogging (N after WL). All other water treatments received 90 kg N/ha at the end of the waterlogging period. These other water treatments were:

• a non-waterlogged (NWL) control that was irrigated to fill the soil profile at the end of the waterlogging period
• a non-irrigated treatment
• an extra irrigation treatment, which was applied after the waterlogging period.

The experimental design consisted of large unreplicated blocks for water treatments. Each block was further divided into three replicates, which contained smaller blocks dedicated to a single crop, to allow easier herbicide application. Canola varieties were Dynatron TT and Nizza CL, barley varieties were RGT-Atlantis and RGT Planet, and faba bean varieties were Amberley and Bendoc. Data were analysed in Genstat^® (24th Edition) by analysis of variance separately for each crop species, with a rep x water treatment stratum, water treatment x N rate stratum, and water treatment x variety x N rate stratum.

Study 3 – Timing of waterlogging and urea application

To test the effect of the timing of both waterlogging and urea application on the growth and yield of canola, barley and faba beans, a field experiment was sown at Hamilton on 25 April 2024. The soil was a Chromosol derived from basalt, with 167 kg/ha of mineral N in the 0–60cm zone and 259 kg N/ha in the 0–100cm zone. Waterlogging timings were early (EWL, July–Aug), mid (MWL, mid-Jul to mid-Sep) and late (LWL, Aug–Sep), which were compared with a non-waterlogged control (NWL) on raised beds that had a soil surface 10–14cm above its adjacent furrows. All canola and barley plots received 10 kg/ha of N as mono-ammonium phosphate (MAP) and 100 kg/ha of N as urea either:

• all broadcast at the start of EWL (N1),
• half coinciding with the start of EWL, half at the end of EWL (N2)
• half coinciding with the start of EWL, half during the EWL period (N3, N4), or
• all at sowing, drilled to 7cm (N5).

A faba bean N treatment compared 50kg N/ha as urea immediately prior to waterlogging with a nil N application. Crop varieties were Pioneer^® 45Y95CL canola, RGT-Atlantis barley and Amberley and Bendoc faba beans.The experiment was laid out with separate sections for each crop, large plots for water treatments and small plots for N timing treatments. Treatment beds were separated by two buffer beds to reduce the amount of water from neighbouring treatments. The maximum rooting depth was assessed by the core break method using two cores of 40mm diameter per plot shortly after the maturity harvest. Data were analysed by analysis of variance in Genstat (24th edition) separately for each crop species, with a rep x water treatment stratum, and a water treatment x N timing stratum.

Results and discussion

Study 1 – Timing and depth of waterlogging

In 2019 and 2021, waterlogging, as indicated by water tables within 30cm of the surface, started in early July and lasted until mid-August at most of the monitoring positions, while water tables persisted in lower positions until early September (Figure 1). In 2021, there were periods when water was above the ground surface (flooding) on the hilltop of the paddock, and nearly a month between mid-July and mid-August, when water tables were within 10cm of the soil surface. By contrast, water tables in 2020 occurred between mid-September and mid-October and were brief.

Figure 1. Timing and depth of perched water tables within 30cm of the surface in Study 1.

The period from early July to mid-August coincides with the period of rapid potential uptake of N and other nutrients. However, spreading of fertiliser can be limited by wet untrafficable soils, and the uptake of N into the plant is limited by anaerobic soil through much of the root zone.

Study 2 – Options for spring sowing and late season waterlogging

Barley was at its anthesis stage by the end of waterlogging and reached maturity on 20 December when it was hand harvested for grain yield. Faba beans flowered from mid-November to mid-December and reached maturity on 19 January. Canola was attacked by cabbage moths and aphids in mid-November and further measurements were discontinued.

At maturity, the biomass of barley was 31–39% lower in waterlogged treatments than the comparable control, except where N had been applied immediately prior to waterlogging (Table 1). By contrast, faba bean maturity biomass was only reduced by 4–20%. The non-irrigated treatment, which provides a measure of crop performance under dry conditions, showed no significant reduction in maturity biomass in barley compared to the control, whereas in faba bean, there was a reduction of 37–42% relative to the control. There were no significant effects of variety (data not shown).

The application of urea immediately before waterlogging was able to maintain the barley crop through the waterlogging period, with no significant reduction in biomass (8.6 vs 7.5 t/ha respectively). An explanation is that, while most of the available N deeper in the soil profile was inaccessible to the crop because these layers were waterlogged, the urea application provided sufficient nutrition in the surface oxidised soil layers for above-ground growth to continue normally.

As a spring sowing option to convert soil water stored from winter into grain, barley proved to be more reliable than faba beans. Its grain-fill was complete by mid-December and therefore, was less sensitive to dry conditions during summer. Faba beans were in their grain-fill period from late November to early January and therefore, were more vulnerable to dry conditions.

Table 1: Maturity biomass (t/ha) of barley and faba beans in Study 2 without added N (0N) and with 90kg/ha of N as urea (90N). Also shown are the p-values for comparison across N treatments within species, the 5% lsd for the water treatment x N interaction, and rainfall plus irrigation from sowing until 2 weeks before harvest. Values are the mean of two varieties.

Study 3 – Timing of waterlogging and urea application

The depth of the water table during the waterlogging period averaged 0cm for barley, 14cm for canola and 4cm for faba beans (Table 2). These depths are relative to the bottom of the drill rows made by the seeder, and there were ridges between planting rows up to 3cm above this depth. At times, when the depth was negative, the plant roots had access to a small volume of non-waterlogged soil. Water tables were deeper in the canola block because it was located uphill of the faba beans, and it was not feasible to lengthen the watering cycle for canola without causing excessive watering of the faba beans. Water also spread laterally into the non-waterlogged control, but the raised beds lifted the soil surface 5–10cm above the water table. Water from surrounding plots extended the period of waterlogging in the barley EWL and MWL treatments beyond the period of irrigation.

Table 2: Average depth (cm below ground) to the perched water table between biomass measurement dates, and the period irrigation was applied to impose waterlogging (shaded).

Barley maturity biomass was reduced by 30–43% relative to the NWL Control (Table 3). The EWL treatment was the most severely affected, as high water tables persisted for 6 weeks after the end of irrigation (Table 2). Canola experienced a 33% reduction in maturity biomass in the EWL treatment, but there was no significant reduction in the MWL and LWL treatments (Table 3). The MWL and LWL treatments for canola had water table depths similar to the NWL controls of the other crop species, meaning that the waterlogging treatments were not as close to the surface as intended. Faba bean maturity biomass was reduced by 51–65% by waterlogging. Water tables in the faba bean block were within 2cm of the surface for most of the waterlogging period and were implemented as intended (Table 2). The faba bean NWL Control had at least 9cm of non-saturated soil, leading to more than double the maturity biomass.

Table 3: Summary of the main effect of water treatment on the maturity biomass of barley, canola and faba beans in Study 3.

The canola results were compromised by water tables that were not as close to the surface as barley or faba bean. A comparable experiment has been undertaken within the same project in Tasmania at a site where water level was controlled, and will provide more detailed data on the effects of shallow water tables on canola growth and yield and results will be presented at the Update.

The N timing treatments showed that banding all N at sowing (N5) was the least effective application method, and that the various split application treatments (N2-N4) were the most effective, but with no significant differences among them (Table 4). The maximum rooting depth at maturity averaged 25cm for barley, 49cm for canola and 32cm for faba beans. Only canola was significantly affected by water treatment, with rooting depths ranging from 35cm for LWL to 63cm for NWL (data not shown). These rooting depths are much shallower than under non-waterlogged conditions, where depths typically exceed 1m. Calculations often assume that mineral N to a depth of 60cm is available to the crop. However, under waterlogged conditions, the zone of N uptake is expected to be restricted to the depth of unsaturated soil, and, after the end of waterlogging, there is limited time for roots to grow to their normal depths.

Applying urea to waterlogged soil carries risks of N loss to leaching, denitrification and ammonia volatilisation. However, the experiments reported here (and previous experiments at Hamilton that were waterlogged) have confirmed that most of the effects of waterlogging on the yield of cereals and canola can be mitigated by broadcasting urea onto wet (but not flooded) soil. Our hypothesis is that, during waterlogging, plant nutrient uptake is confined to the top oxidised soil layers, and that by providing sufficient nutrition in these layers, above-ground crop growth can continue. Further to this hypothesis, the conversion of ammonium to nitrate (nitrification) is delayed in waterlogged soils, as the process requires oxygen, leading to most of the applied N remaining in the ammonium form until taken up by plants. The ammonium form leaches much more slowly than nitrate and does not denitrify until its conversion to nitrate, but can be lost to ammonia volatilisation. The banded urea may have had sufficient time to nitrify prior to the start of waterlogging, whereas the split applications were applied closer to the time of plant demand and so, were less vulnerable to loss.  Keys to minimising losses are

• not applying urea when soil is flooded as much of it will move off the paddock
• minimising the time between urea application and the start of waterlogging, to ensure as much as possible is retained in the ammonium form
• apply only a portion of the N requirements prior to waterlogging, with the remainder during or after waterlogging
• maintaining high yields, as emissions intensity in greenhouse gas accounting is calculated per tonne of product.

Table 4: Dates of application for each N timing treatment (N1 to N5), and the maturity biomass (t/ha) of barley and canola. Also shown is the 5% Lsd for comparisons within water treatment (rows), and within N treatment (columns).

Conclusion

Some of the effects of waterlogging on crop growth can be mitigated by broadcasting urea onto saturated (but not flooded) soil. This is consistent with the hypothesis that nutrient uptake by HRZ crops is confined to a thin oxidised layer of surface soil. Since root growth extension to deeper depths is also delayed, soil nitrogen deeper than 30cm should be considered unavailable to crops that have experienced waterlogging.

Acknowledgements

The research undertaken as part of this project is made possible by the significant contributions of growers through both trial cooperation and the support of the GRDC, the authors would like to thank them for their continued support. We thank our technical staff for their skilful implementation of these experiments, and the wider team of the Waterlogging project for their input.

References

Robertson F, Suraweera D, McCaskill M, Christy B, Armstrong R, Zollinger R, Byron J, Partington D, Clark S (2019) Waterlogging effects on soils and wheat crops in the high rainfall zone of Victoria. Proceedings of the 19th Australian Society of Agronomy Conference, Wagga Wagga, NSW, Australia, 25–29 August 2019. (http://www.agronomyaustraliaproceedings.org/images/sampledata/2019/2019ASA_Robertson_Fiona_339.pdf).

Contact details

Malcolm McCaskill
Agriculture Victoria Research
915 Mount Napier Road, Hamilton VIC 330
0407 850 671
malcolm.mccaskill@agriculture.vic.gov.au