Frost response in lentils

Take home messages

• The Mallee experienced an above average number of frost days during 2018, where the trial site (12km north west of Ouyen) recorded 37 mornings below 0⁰C between August and October.
• Lentils were most susceptible to frost during the flowering and pod filling stage, where temperatures below 0⁰C at canopy height decrease grain yield and quality.
• Frost applied at flowering reduced grain yield when a threshold of 31°C.hr (<0°C) was reached, with a yield decline of 3.8% per ⁰C.hr.
• Frost applied during pod filling reduced grain yield by 2% per ⁰C (<0⁰C).
• Under severe frost conditions experienced in 2018, conventional and imidazolinone (imi) tolerant lentils were equally affected by natural frosts, indicating that imi tolerance was not linked to increased frost sensitivity.

Background

Radiant frost can significantly reduce the yield and quality of lentils, with economic losses due to frost damage in broadacre cropping estimated to be $360 million per year in Australia (March et al.2015, Watt. 2013, Rebbeck et al. 2007). For lentils, frost damage can occur at any development stage following emergence, but the greatest potential for damage coincides with the reproductive period.

Over recent years, frost damage has occurred across widespread parts of the Mallee and Wimmera and caused significant reductions in grain yield and quality. Recent increases in the severity and duration of frost across southern Australia (Crimp et al. 2016), combined with the widespread adoption of earlier sowing of pulses, have increased crop exposure to frost damage. Furthermore, industry perception exists that Group B herbicide tolerant lentil varieties suffer greater yield reductions compared to conventional lentil varieties because of frost.

Management strategies to mitigate frost damage are through avoidance, manipulating variables such as sowing date, crop and variety selections. However, these strategies can create alternative problems, including shifting the reproductive phase further into the heatwave window. Reducing stubble loads also can reduce the severity of frost, however, this does not offer a practical solution where yield gains associated with standing stubble are greater.

This paper reports on field trials conducted at Horsham in 2017 and Ouyen in 2018 which aimed to identify the fundamental response of lentils to frost damage, opportunities to detect damage using electronic sensors and the relative response of Group B tolerant varieties compared with conventional varieties.

Method

Mobile frost chambers were used to examine the effect of simulated frost on lentil growth and yield near Horsham (2017) and Ouyen, Victoria (2018). Experimental work in 2017 assessed the response of lentils (cv. PBA Jumbo 2) to 12 frost scenarios, where temperatures below 0°C were applied at flowering, early pod, flat pod, and pre-filled pod. In 2018, experimental work at Ouyen was designed to determine if Group B herbicide (imi)tolerance in lentils is linked to increased sensitivity to frost.  In this trial, frost was applied at either the late vegetative or late podding stage. The trial was replicated six times and included six lentil varieties differing in imi rating.

Imi tolerant varieties — PBA Hurricane XT, PBA Herald XT, PBA Hallmark XT, CIPAL 1721.

Conventional varieties — PBA Jumbo 2, PBA Flash.

Frost application

In both years, temperatures below 0°C were applied to lentils, using mobile field frost chambers (Nuttall et al. 2018), where these simulated frost events were applied over a single night between 9pm and midnight. These were compared with two sets of control plots — open control (OC) constituting lentils grown under ambient air throughout the growing period, and a chamber control (CC) where plants were protected from frost using chambers which were installed when frost conditions were forecast. During the 2018 season, the OC plots included natural frosts.

Cold load

To account for the varying severity in frost (temperature × duration) imposed on lentils, the cold load was calculated as the sum of degrees Celsius (⁰C) below 0⁰C for the logged temperature data to give a ⁰C.hr.

Remote sensing

Proximal sensing was used to monitor the crop to determine whether non-destructive methods could be used to detect early (pre-visual) frost damage within lentils. A portable spectrometer was used to determine reflectance indices including normalised difference vegetation index (NDVI) (Rouse Jr, 1974 #682) and photochemical response index (PRI) (Gamon, 1992 #636).

Results and discussion

2018 season

In 2018, the Mallee sustained a significant number of frost days during late winter and spring, coinciding with late vegetative and early reproductive phase of many crops. The trial site 12km north west of Ouyen recorded 37 mornings below 0⁰C at canopy height between August and October with temperatures dropping below -2⁰C on 26 of these mornings. The crop canopy temperature was approximately 2⁰C lower than that recorded at 1.2m above the ground surface in a Stevenson screen (1.2m) (data not shown). This demonstrates that crops experience colder temperatures than those generally recorded by weather stations. Growing season rainfall (April-October inclusive) was 93mm, well below the long-term average of 178mm.

Figure 1. Cumulative cold load ⁰C.hr (<0⁰C) of natural frost recorded received by lentils during the reproductive period at the Ouyen trial site during 2018. Cold load ⁰C.hr (<0⁰C) is the total time and temperature crops were exposed to below zero degrees at canopy.

Timing of frost

A relationship between lentil yield response and cold load was defined using field data during 2017 at Horsham, Victoria, where no natural frost was recorded during the reproductive window. At flowering, damage occurs when a threshold of 31⁰C.hr (<0⁰C) is reached, hereafter yield decline was 3.8% per ⁰C.h. At pod filling for every degree hour below zero, there was a 2% reduction in grain yield. The difference in response to frost at flowering and pod filling indicates that timing, intensity and duration affect the extent to which lentils recover from frost. The capacity for lentils to recover is greater at flowering, but limited by intensity and duration.

The response of lentils to frost in 2018 differed to that observed in 2017 (Figure 2). In 2018, absolute yields were lower and there was no clear relationship between cold load (applied and natural) and grain yield. This is due to a combination of dry seasonal conditions (decile 1) and multiple, severe natural frosts during the reproductive period. For the 2018 trial, the OCs (crops that were exposed to natural frosts only) received 78⁰C.hr (<0⁰C), during the treatment window (five days pre and post vegetative and pod filling frost application) compared to the maximum cold load applied to the treated plots in 2017 (67⁰C.hr <0⁰C). Yields measured in 2018 suggest that in years where yield potential is lower, the effect of frost is less compared to years with average rainfall.

Figure 2. Relationship between lentil cv. PBA Jumbo 2 grain yield and cold load (°C.hr <0°C) associated with frost treatments applied at five different growth stages — late vegetative, flowering, early pod, flat pod and filling pod. The 2017 data (black) is fitted by two regression models. The 2017 CC and flowering data were fitted with a segmented regression (dash line). The 2017 CC, OC and pod filling stages were fitted with a linear regression (solid line). The 2018 data (grey) have cold loads >70 ⁰C.hr <0⁰C, and cold loads did not relate to yield reduction. For this analysis, the cold loads were defined by the cumulative temperature below 0⁰C during the treatment window (five days either side of applied treatments), when the canopy was exposed to natural and applied frost.

Lentil marketability is strongly influenced by visual characteristics such as discolouration, deformation and shrivelling. Frost, as well as other abiotic stresses, can affect these qualities. For lentils exposed to a range of frost scenarios (2017 trials), there was a corresponding degradation in visual quality with increasing cold load (Figure 3). In contrast, there was minimal damage to lentil grain when frost was applied at flowering, regardless of intensity and duration (data not shown).

Figure 3. Frost affected lentil grain. Visual characteristics of lentil cv. Jumbo 2 where frost treatments of varying intensities have been imposed on plants during the pod filling phase. A control (5⁰C.hr <0⁰C) a) is compared with b) 16, c) 22 and d) 43⁰C.h.

Variety response to frost

Figure 4. Comparison of conventional to imi varieties tested in 2018 field trial at Ouyen. The differences in grain yield across conventional and imi varieties a), and the reduction in grain yield (%) of frost applied at vegetative and pod filling, compared to natural frost only b). Frost treatments are natural frost, applied at vegetative and natural frost, applied at reproductive and natural frost.

For the natural frost conditions which occurred in 2018, all varieties were equally susceptible to frost (p = 0.441), indicating that imi tolerance was not linked to increased frost sensitivity under these conditions (Figure 4a). Furthermore, the cumulative yield impact induced by the artificial frost applied at the late vegetative and podding stage, over the natural frost effects, was also equivalent for the conventional and imi varieties (Figure 4b). Future work would include verifying this pattern of response to frost under wetter growing conditions where yield potential is greater than in 2018, and alternative frost stress patterns.

For pooled response across conventional and imi varieties, there was an overall reduction in grain yield of 40% and 13% for artificial frost pyramided on natural frosts (average minimum -2.4⁰C), applied during the vegetative (average minimum -5.07⁰C) and reproductive (average minimum -4.58⁰C) phase, respectively (Figure 4a). The effect of the applied frost was equivalent for the late vegetative and pod filling treatments.

Early detection using remote sensing

The impact of frost to lentils was measured using proximal sensing following the application of artificial frosts in 2017. Several reflectance indices were strongly correlated with cold load, including NDVI and PRI. There was a strong negative correlation between cold load and NDVI where cold loads exceeded 31.3⁰C.hr (<0⁰C). PRI was more sensitive to frost damage compared to NDVI, where beyond a threshold of 22.39 °C.hr (<0⁰C), PRI increased, indicating a step-change response induced by frost above this threshold.

Figure 5. Reflectance indices for lentil cv. Jumbo 2 exposed to frost at varying intensities expressed as cold loads. Reflectance indices a) NDVI and b) PRI were measured using a portable spectrometer 10cm above the canopy at 6 days after frost (DAFr)and 13 DAFr (following flowering application) and 3 DAFr (following podding treatment). Segmented regression models are for CC, OC, flowering and early podding data. 0 DAFr is the day following the night of frost application.

The good agreement of NDVI and PRI to crop reflectance of frost affected crops supports the potential to utilise non-destructive measurements using proximal and remote sensing tools (e.g. vehicle mounted, airborne or satellite imagery). This work requires ongoing validation to confirm the utility of these and other potential indices and extend the assessment of frost damage to a larger scale spatial context (e.g. paddock scale).

Conclusion

Radiant frost continues to limit production of lentils in southern Australia, reducing grain yield and causing deformation of grain. Lentils are susceptible to frost damage at any time from emergence to maturity, but are most susceptible to frost during the pod filling stage, where we defined for every degree hour below zero, there is a 2% reduction in grain yield. There was no difference between imi tolerant and conventional varieties, under severe frost conditions, from the late vegetative to late pod filling stage. Preliminary work shows that there is a strong relationship between NDVI and PRI with cold load, indicating that there is potential to apply these technologies on a paddock scale to assess frost damage. Ongoing work is required to build on the fundamental response of lentils to frost effects, where the influence of the indeterminate growth habit, imi tolerance and diagnostics using remote sensing need further defining.

References

Crimp S, Gobbett, D Kokic, P Nidumolu, U, Howden M, Nicholls N (2016). Recent seasonal and long-term changes in southern Australian frost occurrence. Climatic Change. 139(1), 115 -128.

Gamon J. A Peñuelas J, Field, CB (1992). A narrow- waveband spectral index that tracks diurnal changes in photosynthetic efficiency. Remote Sensing of Environment. 41, 35-44.

March T, Knights S, Biddulph B, Ogbonnaya F, Maccallum R, Belford R (2015). The GRDC National Frost Initiative, GRDC Update paper (https://grdc.com.au/resources-and-publications/grdc-update-papers/tab-content/grdc-update-papers/2015/02/the-grdc-national-frost-initiative) .

Nuttall JG, Perry EM, Delahunty AJ, O’Leary GJ, Barlow KM, Wallace AJ (2018). Frost response on wheat and early detection using proximal sensors. Journal of Agronomy and Crop Science, 1-15.

Rouse J. Jr, Haas R, Schell J, Deering D (1974). Monitoring vegetation systems in the Great Plains with ERTS.

Rebbeck MA, Knell GR, Hayman P T, Lynch CW, Alexander, BM, Faulkner M, Falconer D (2007). Agronomic practices to reduce frost risk. In D. Reuter (Ed.), Managing frost risk – A guide for southern Australia grains.Canberra, ACT: South Australian Research and Development Institute and Grains Research and Development Corporation.

Acknowledgments

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 author would like to thank them for their continued support.

This work was funded by the ‘Improving practices and adoption through strengthening D&E capability and delivery in the southern region’, Regional Research Agronomist program (DAV00143) as part of the GRDC and Department of Jobs, Precincts and Regions Agreement. The authors also wish to acknowledge the assistance of Moodie Agronomy staff in delivery of the experimental program. Assistance from Agriculture Victoria staff Ashley Purdue, Kate Finger, Mitchell Fromm and Alexander Clancy was also paramount to carrying out this work.

Contact

Audrey Delahunty
Agriculture Victoria
Cnr Eleventh St & Koorlong Ave, Irymple, Vic, 3498
0427 580 131
audrey.j.delahunty@ecodev.vic.gov.au
@AudreyDelahunty