Trinexapac-ethyl delays lodging and increases seed yield in perennial ryegrass seed crops

Abstract

The plant growth regulator trinexapac-ethyl (TE) is widely used to enhance ryegrass seed yields by reducing lodging (i.e. delaying collapse of the crop). However, lodging data are usually a single evaluation related to TE rate for a particular date of measurement. In eight field trials with varying rates of TE, weekly lodging assessments were made from full head emergence and days to 50% lodging were determined. In all trials, dynamic assessment of TE treatment showed there was a strong positive correlation (R ² = 0.82) between the number of days to 50% lodging and seed yield. Nil TE reached 50% lodging in 7 days compared with 33 days for TE treatments of 400 g ha⁻¹. On average, one days’ delay in reaching 50% lodging increased seed yield by 24 kg ha⁻¹ (response range 16–33 kg ha⁻¹). The data indicate that TE rate per se is less important than the rate required to delay the date that 50% lodging is reached. The seed yield response was linear for delayed lodging from full head emergence to harvest, suggesting that the highest yields are achieved if the crop is not more than 50% lodged at harvest.

Keywords:

• lodging
• Lolium perenne
• perennial ryegrass
• plant growth regulator
• seed yield
• trinexapac-ethyl

Introduction

The plant growth regulator trinexapac-ethyl (TE) is used by 95% of ryegrass seed growers in New Zealand (Rolston et al. 2004; Chynoweth et al. 2010). TE acts as a stem shortener, increasing stem strength and reducing lodging (Chastain et al. 2003) by inhibiting gibberellic acid biosynthesis (Rademacher 2000). Lodging is the collapse of the crop, usually at or after flowering, when the seed head can no longer be supported by the stem. The optimum time of TE application is at the appearance of the second node, described as Zadoks Growth Stage (GS) 32 (Zadoks et al. 1974). Seed yield increases in perennial ryegrass (Lolium perenne L.) from TE rates up to 300–400 g ha⁻¹ are typically 30–50% and the rate response to TE is often linear to rates up to of 800 g ha⁻¹ (Rolston et al. 2004). A survey by the authors of 22 New Zealand growers based on their crop dairy records in 2007/08 showed TE average use rates of 330 g ha⁻¹, with a range of 250–400 g ha⁻¹. Growers are reluctant to use higher rates of TE (>500 g ha⁻¹) because they fear that crops that do not lodge by harvest will be vulnerable to seed shattering in windy conditions (personal communication, Murray Kelly, PGG Wrightson Seeds). In previous reported trials on TE response, lodging data are usually from a single evaluation and lodging is related to TE rate for the particular date of measurement (Rolston et al. 2005; Macháč 2010). Borm & van den Berg (2008) made weekly lodging assessments and concluded the TE delayed lodging but seed yield responses were small 6—13%. In the present study trials were undertaken in four harvest seasons to determine both TE rate response and to understand variation in rate response that we observed in growers fields. Dynamic assessments of lodging were recorded weekly and the change in lodging determined over a period of seven weeks from full head emergence through anthesis and seed fill. Days to 50% lodging was calculated and the relationship to seed yield assessed.

Method

Eight TE rate and timing response trials were undertaken in ryegrass seed production fields in the Canterbury region of New Zealand during four seasons from 2006/07 to 2009/10. The trials included a range of cultivars and locations (Table 1). All inputs except TE, and nitrogen (N) in Trial 1, were managed by the grower; TE applications were applied by the research team. Trials had four replicates in a randomised block design. Plots were 3.2 m wide and at least 9 m long. TE (Moddus® 250 g l⁻¹ TE in the form of an emulsifiable concentrate) was applied with a plot sprayer at 220 l water ha⁻¹. Each trial consisted of between three and five rates of TE applied to different N rates (Trial 1), different closing dates (Trials 2, 4 and 5) or as single versus split rates of TE or combinations of TE and CCC (chlormequat-chloride) (Trials 3, 6, 7 and 8). The trials each had between 8 and 15 treatments. Weekly assessments of lodging percentage were taken with lodging been assessed on a 0 (no lodging) to 100% (fully lodged) linear scale. Days to 50% lodging from full head emergence, Zadoks GS 59, was calculated. Components of seed yield were assessed from 0.25 m² quadrats cut at late seed fill.

Table 1  Location, year, cultivar, range of TE rates evaluated, irrigation used, seed yield for two rates of TE and increase in seed yield from TE 400 g ha⁻¹ and nil TE.

						Seed yield (kg ha^−1)
Trial	Location	Year	Cultivar	TE rate (g ha^−1)	Irrigated	TE 0	TE 400	LSD 5%	Increase in seed yield (%)
1	Wakanui	06/07	Hillary	0–400	Yes	1320	2240	202	69
2	Methven	07/08	Commando	0–800	No	1650	2340	251	42
3	Dunsandel	08/09	Extreme	0–600	Yes	1180	1710	123	45
4	Lincoln	08/09	Commando	0–400	Yes	1980	2840	270	43
5	Lincoln	08/09	Nui	0–400	Yes	1800	2580	270	43
6	Wakanui	08/09	Bealey	0–600	Yes	2380	2960	329	24
7	Wakanui	09/10	Bealey	0–600	Yes	2240	2880	279	28
8	Dunsandel	09/10	Alto	0–600	Yes	1020	1590	175	56
Average						1700	2400		44

At harvest, a 1.7 m swath was cut from the centre of all plots with a modified plot windrower, and then harvested with a ‘Sampo’ plot combine. The crop was windrowed on the day the farmer cut his field at ca. 40% seed moisture content (SMC) and combined about 7 days later when seed had dried to <14% SMC. Seed samples were machine dressed on a small-scale air-screen separator to achieve a first-generation seed purity standard (MAF 2008). Cleaned plot samples were weighed and converted to yield per hectare. GenStat (version 10) was used for statistical analysis using a general ANOVA (analysis of variance) model. Significant interactions and effects were separated using least significant difference (LSD) tests (α = 0.05).

Results and discussion

Over eight trials, the average seed yield response to 400 g ha⁻¹ TE compared to nil was 44%, increasing seed yield by 700 kg ha⁻¹ from an average of 1700 kg ha⁻¹ in the untreated control (Table 1). The seed yield response is in the range reported by Rolston et al. (2004 ,2005). Dynamic assessment of weekly lodging scores showed that the average days to 50% lodging in the nil TE treatments was 7 days, compared with 33 days for TE 400 g ha⁻¹ (Table 2). With no TE, seed fill (the period from anthesis to seed being physiologically mature) occurs when the plant is in a lodged state; at the TE 400 rate, plants can complete seed fill in either an erect or semi-erect state. The increase in seed yield with delayed days to 50% lodging was linear with an average correlation coefficient of R ²=0.81 (Table 2). The data from Trial 4 are presented as a typical example of the six trials in Fig. 1. The average seed yield response to delaying days to 50% lodging was 24 kg ha⁻¹ for each day's delay (Table 2). Seed yield response in these trials was the result of a higher conversion of fertile florets to saleable seed, resulting in more seeds per unit area (Table 3). The TE treatment had no effect on seed head density, spikelet numbers per head, florets per spikelet or seed weight (expressed as thousand seed weight).

Fig. 1  Relationship between seed yield and days to 50% lodging for eight TE treatments (Trial 4).

Table 2  Seed yield increase for each day's delay to 50% lodging, standard error of the mean (SEM) and correlation coefficient (R ²) for each linear regression and days to 50% lodging from full head emergence for nil and 400 g ha⁻¹ TE treatment for eight trials.

			Days 50% lodging
Trial	Seed yield increase (kg ha^−1 day^−1)	R ^2	TE 0	TE 400
1	33	0.93	3±1	33±1
2	19	0.80	1±1	32±1
3	16	0.95	10±1	42±2
4	26	0.84	9±1	34±6
5	20	0.59	9±1	43±4
6	30	0.69	10±2	24±2
7	32	0.91	12±2	26±3
8	16	0.82	1±1	32±3
Average	24	0.81	7	33
SEM	±2	±0.05	±2	±3

Table 3  Harvest mass, reproductive stem length and components of seed yield (Trial 3)

Component	TE 0	TE 400	LSD 5%
Harvest mass (t ha^−1)	16.1	14.1	NS*
Stem length (mm)	870	640	55
Heads (m^2)	2180	2190	NS
Spikelets (head)	19.0	19.3	NS
Fertile florets (spikelet)	4.1	4.4	NS
Sterile florets (spikelet)	1.1	1.2	NS
Total floret (spikelet)	5.2	5.6	NS
TSW (g)	2.86	2.79	NS
Seed yield (kg ha^−1)	1180	1710	123
Seeds (000 m^−2)	41.4	61.4	14.8
Seeds per spikelet	1.1	1.8	0.5
Harvest index (%)	7.3	12.1	1.5
*not significant

Early lodged crops often trigger secondary vegetative tillering, which is postulated as either resulting in nutrient competition with seed fill or reducing the level of photosynthetic radiation to the crop (Rolston et al. 2007). Early lodging may also reduce the amount of photosynthetic radiation within the crop (Trethewey and Rolston 2009) and further work on this relationship is required.

Conclusion

The plant growth regulator TE increases ryegrass seed yields by delaying the onset of lodging. The number of days from full head emergence to when the crop is 50% lodged is highly correlated with seed yield. Seed growers are recommended to use rates of TE to achieve a crop that is ≤50% lodged at harvest.

Acknowledgements

The authors wish to thank the four growers (Jonathon Clay, Eric Watson, Graham Lill and Roger Orchard AgResearch Lincoln farm manager) who made land available for the trials. The Foundation for Arable Research (FAR) funded the study.