Abstract

Key words

Root failure, lodging, wheat, barley, image analysis

Introduction

The process by which shoots of cereals are displaced from their vertical orientation is known as lodging (Pinthus 1973). Lodging is most likely during the two or three months preceding harvest, usually after ear or panicle emergence with the result that shoots permanently lean or lie horizontally on the ground. Two forms of lodging have been recognised: stem buckling (stem lodging) (Thomas 1982) or displacement of roots within the soil (root lodging) (Ennos 1991). In stem lodging, roots are held firmly in a strong soil where the wind force buckles one of the lower internodes of the shoot. Root lodging becomes more likely when the anchorage strength is reduced by weak soil or poorly developed anchorage roots. The resulting effect is a reduction in crop yield by up to 80%, with further losses in grain quality, greater drying costs and an increase in time taken for harvesting. Lodging is not a new agricultural phenomenon, but is a problem that limits cereal productivity in both the developed and developing world (Berry et al. 2004).

In wheat, barley and oats, stem lodging is usually caused by one of the bottom two internodes buckling (Neenan and Spencer-Smith 1975; Mulder 1954) and results in the upper stem and ear lying horizontally. Buckling of the middle internodes is more common in barley (Neenan and Spencer-Smith 1975) and oats (White et al. 2003). Root lodging has been observed in wheat (Crook and Ennos 1993; Easson et al. 1993), barley (Graham 1983) and oats (Mulder 1954). The process of root lodging results in the permanent displacement of cereal stems without any observable stem buckling. A possible outcome of this is that plants lean at less than 90° from the vertical or lie horizontally. Lodging events themselves have been rarely observed in any scientific way (Sterling et al. 2003), which is simply due to the unpredictable nature of lodging in both spatial and temporal terms, together with the difficulty of making field measurements in often hostile weather conditions (Sterling et al. 2003). Above ground observations of lodging have been modelled by Baker et al. (1998), Berry et al. (2000, 2002, 2003) and Sterling et al. (2003). However, these were only above-ground observations and no suggestion was made concerning the actual lodging process below ground. Ennos (1991) produced the first below-ground observation of the lodging process using spring wheat, with what he termed the ‘Root Sliding Model’. Here, as the plant is pushed over, the root system rotates about a point 10-15 mm below the base of the stem, the coronal roots are bent at their bases and they move axially into and out of the soil. Further observations were made by Ennos et al. (1993) who suggested a ‘Hinge Model’, based on the contrasting anchorage characteristics of Himalayan Balsam and Mature Sunflower. Here, observations were made where once the stem has been pulled over by the wind, the plants rotated about a hinge on the leeward side of the stem, which results in a root-soil ball being pulled out of the soil. A third model, suggested by Crook and Ennos (1993) using the anchorage mechanics of mature winter wheat, produced the ‘Root Cone Model’. This form of anchorage failure occurs when the root-soil cone is rotated at its windward edge, forcing the soil inside the cone to push further down into the soil as a block, thus compressing the soil below it. As the soil is pushed past its plastic limit it becomes deformed and does not return to its original form.

Crook and Ennos (1993) argued that root lodging should be predominant in modern wheat cultivars, whereas Neenan and Spencer-Smith (1975) favoured stem lodging as the main failure mechanism. Baker et al. (1998) illustrated that both types of lodging were possible depending on the particular crop characteristics. Sterling et al., (2003) confirmed this by direct observations of both lodging mechanisms during wind tunnel experiments on field grown winter wheat. Berry et al. (2003) stated that the most likely form of lodging depends upon the environment at the time of lodging and during the growth of the crop. For example, wet and weak soil increases the risk of root lodging over the risk of stem lodging, whereas high levels of soil mineral nitrogen increase the risk of stem lodging more than the risk of root lodging.

There are many factors that contribute to the process of lodging, the most predominant of these include: wind, rain, topography, soil type, crop husbandry practices, crop disease and an abundant supply of nutrients in the soil (Berry et al. 2004). Attempts at reducing the lodging risk of crops were introduced in the 1960’s and 1970’s with the production of semi-dwarf plant species, the resulting effect being an increase in crop grain yields. Plant growth regulators (PGRs) have also been used to decrease crop height and further reduce the lodging risk of cereals. Three major types of PGRs have been introduced including: chlormequat chloride (1960s), ethepon (late 1980s) and trinexapac-ethyl (mid 1990s). In France, Germany and the UK, which have among the largest cereal yields in the world, PGR’s are now applied to more than 70% of wheat area (Berry et al. 2004). A lower plant density has been shown to reduce lodging. Berry et al. (2000) found that by establishing 200 wheat plants m^-2 compared with 400 wheat plants m^-2 reduced the lodging risk by increasing the strength of the anchorage system by more than 50% and the strength of the stem by 15%. The state of the soil has also been found to have an important effect on lodging, as it has been predicted to be directly proportional to anchorage strength of the crop. A model of soil strength developed by Baker et al. (1998) showed that variation in clay content, moisture content and compaction could each be expected to alter the soil shear strength by several fold. However, this model has not been widely tested on varying soil types and does not account for possible interactions between the soil factors and therefore the results from the model need to be treated with caution.

This paper explores the in situ observation of root lodging in winter wheat and barley crops grown at two plant densities (100 and 400 seeds/m²) on three different soil textures (Sandy Loam, Silty Loam and Clay) in the UK. Observations were made using image analysis to quantify differences in the size, shape and distribution of soil pores.

Materials and Methods

Soils and crops

Soil samples were collected from a sandy loam of the Dunnington Heath series (FAO: Chromi-Abruptic Luvisol) at the University of Nottingham’s experimental Farm Site, Sutton Bonington, Nottinghamshire, UK (GR SK504262), a silty loam of the Bromyard Series (FAO: Chromic Luvisol) at ADAS Rosemaund, Herefordshire, UK (GR SO355246) and a clay of the Hanslope Series (FAO: Calcaric-Endostagnic Cambisol) at ADAS Boxworth, Cambridgeshire, UK (GR TL534264). Selected physical characteristics of these soils are shown in Table 1. The soils were sown with winter wheat (Triticum aestivum L.) and winter barley (Hordeum vulgare L.) at 100 and 400 seeds m^-2. Selected root characteristics are illustrated in Table 2. Cultivations at the time of sowing included inversion tillage to 20 cm depth. Trials were set in randomized blocks, with three blocks each of the two sowing densities for both wheat and barley.

Table 1: Physical characteristics of the three soils.

Soil Series	Soil Texture	Bulk Density (g cm –3)	Organic content (%)
Dunnington Heath (Sutton Bonington, Nottinghamshire, UK)	Sandy Loam (66% sand, 18% silt, 16% clay)	1.42	1.2
Bromyard (ADAS Rosemaund, Herefordshire, UK)	Silt Loam (65% silt, 25% clay, 10% sand)	1.24	5.4
Hanslope (ADAS Boxworth, Cambridgeshire, UK)	Clay (34.7% clay, 34.8% silt, 30.4% sand)	1.21	1.6

Table 2: Root plate spread (mm) and number of roots at each plant density in wheat and barley for each site.

	Root plate spread (mm)				Crown root number per plant
	Clay	Silty loam	Sandy loam	Mean	Clay	Silty loam	Sandy loam	Mean
Wheat
100 seed m2	59.3	48.5	47.7	51.8	48.4	32.9	24.7	35.3
400 seed m2	37.2	37.9	35.7	36.9	21.0	18.4	13.0	17.5
Mean	48.3	43.2	41.7	44.4	34.7	25.7	18.9	26.4
Barley
100 seed m2	48.7	28.0	39.0	38.6	35.2	24.3	26.4	28.6
400 seed m2	25.8	21.8	28.7	25.4	17.8	16.5	14.9	16.4
Mean	37.3	24.9	33.9	32.0	26.5	20.4	20.7	22.5
Overall Mean	42.8	39.1	37.8	39.9	30.6	23.1	19.8	24.5
Site P-Value	0.015				<0.001
Site SED (24 df)	2.78				2.45
Species P-Value	<0.001				0.065
Species SED (24 df)	2.27				2.00
Seed rate P-Value	<0.001				<0.001
Seed rate SED (24 df)	2.27				2.00

Collection of field samples

Prior to lodging, soils were irrigated at a rate of 10 mm hr^-1, bringing the soil to near field capacity in order to promote cereal anchorage failure. Circular steel collars (120 mm diameter) were gently inserted into the ground to 10 mm depth, around the base of the plant stem. 250 ml of household varnish (Tams et al. 2004) was slowly applied to the surface of the soil in 50 ml pulses. Once the initial impregnation had fully infiltrated the soil, the remaining varnish was added in further 50 ml increments (Figure 1) and left to cure for 7-14 days. After this time, samples were carefully dug out of the ground, packed into containers and transported as consolidated blocks to the laboratory. Three lodged plants and three unlodged plants were collected from each plot, giving a total of 72 samples per soil type and 216 samples across all three soil types.

Figure 1: Illustration of a lodged sample following field impregnation with varnish.

Sample preparation

Samples were transferred to 1728 cm³ containers for a secondary impregnation with a 1L mixture of ratio 50:50 acetone and Crystic resin (17749), 10 ml Catalyst ‘O’, 1 ml Accelerator G (Aeropia Ltd., UK) and 0.5 g fluorescent dye (Ciba Geigy, UK). Samples were left for 24 h in a fume cupboard then ‘topped up’ for an average of three to four days until the level of the impregnating mixture did not fall below the surface of the sample. Acetone was omitted from the final mixture, with the volume required dependent on the amount of evaporated acetone and level of drop from the sample surface. Samples were then cured in a fume cupboard for three weeks before being incubated at 25°C for 14 days then 40°C for 7 days to harden. The solid blocks were cut into vertical slices using an oil-lubricated diamond saw and the resultant blocks polished in preparation for digital image acquisition.

Image acquisition and analysis

High-resolution (1600 x 1200 pixels, spatial resolution 90 µm pixel^-1) digital images were collected in UV light (Figure 2(a)) using a TIFF file format and transferred to a PC. The images were processed using analySIS 3.0 auto version (SIS Munster, Germany). The first stage of analysis involved thresholding the image, which defined limits based on the histogram of the red, green and blue colour values of the pixels within an image to segment the soil and the pore space. Images were then binarised (black and white) (Figure 2(b)), subjected to a 1 pixel erosion and pore characteristics measured (total porosity, mean pore area, mean pore shape, mean pore sphericity) (Figure 2(c) colour indicates pore spaces identified for measurement). These attributed were measured using algorithms in the analySIS 3.0 program. The shape factor (the ‘roundness’ of the pore) was classified in accordance with Mooney et al. (2000), where 0.0 – 0.6 = elongate, 0.6 – 0.8 = irregular and 0.8 – 1.0 = rounded.

Figure 2: Stages of image capture and analysis illustrated by an example image of a low density wheat plant in clay soil ((a) digital image before processing, (b) image in binarised form and (c) measurement of pore characteristics).

Results and Discussion

Figure 3: Mean total macroporosity measurements at high and low density for lodged (L) /unlodged (UL) plants in the three soil types.

It was hypothesised that lodging would reduce the total porosity of the soils due to compaction of pores induced by the movement of the root-soil complex. From the results (Figure 3) it was shown that lodging had a significant effect (p <0.001) on the total porosity of the three soil types. Across all treatments, lodging reduced soil porosity from 17.7 to 13.9% (p<0.001). There was an interaction between lodging and crop species (p<0.05) because lodging reduced porosity more in barley sown plots (20.2 to 14.1%) than in wheat sown plots (15.3% to 13.7%). There was a slight interaction between lodging and soil type (p=0.075) because lodging reduced porosity more on the sandy loam than on the clay or silty loam. The greatest effect of lodging on total porosity in the high density treatments was found in the silty loam sown to wheat, where total porosity increased by an average of 5.0%. In both the clay and sandy loam, lodging reduced porosity by an average of 3.7% and 1.7% in the high density treatments. For lodged plants in the low density wheat treatments, total porosity increased by 3.38% in the clay with both the silty and sandy loams producing slight decreases of 0.2% and 0.2% respectively. In the barley treatments, at high density both the clay and sandy loam soils had a reduction in porosity due to lodging, the greatest in the clay (8.8%) compared to the sandy loam (0.7%). The silty loam had an increase of porosity by 1.3%. The low density treatments had similar patterns to the high density treatments, where lodging reduced the total porosity of both the clay and sandy loam soils. The greatest reduction occurred in the sandy loam (5.5%) compared to the 1.5% reduction in the clay. Similarly to the high density barley, the low density treatment for the silty loam had an increase of porosity, however this was marginal at 0.3%.

The results indicated that lodging appeared to have had the greatest effect on the clay soil where changes in porosity were most prevalent in both crop types at both densities. In three out of the four treatments for this soil, total porosity was reduced (wheat high, barley high and low), but increased in the low density wheat treatment. It is possible that these results are due to the fine textured nature of the clay, where pores are more easily compacted. A further possible explanation for these results is concerned with the root characteristics of the plants grown in the clay soil. For the high density wheat and high/low barley treatments, the root plate spread and number of roots were lowest in these treatments compared to the low density seed rate wheat treatment. This reduced number of roots and lower root plate spread (Table 2) would mean that there is more soil available for pore reduction to occur, but at the same time there are also fewer roots to affect the pore reduction. In the low density treatment the number and spread of roots is increased perhaps indicating that a greater root mass increased the lodging movement and increased pore space. Alternatively, it could indicate that these roots also grew into the pre-existing pore space.

Figure 4: Mean pore shape for each of the soil and crop types for lodged (L)/unlodged (UL) treatments.

It was expected that lodging would induce pore changes through the greater elongation of the pores. However, as indicated in Figure 4, there was little significant difference between lodged and unlodged pores and between crop types in all three soil types, with the mean pore shape lying between 0.6 – 0.7 revealing that the majority of pores in all three soils remained irregularly shaped after lodging. The clay soil was an exception, where in the high density wheat there was a clear difference between the unlodged and lodged treatments. In the unlodged clay soil the mean pore shape was rounded, but following lodging, the mean pore shape was re-classified as irregular indicating that the lodging process has had some impact on the pore component of the clay. This result was not surprising given the fine texture of the clay, however it would be expected that any changes in pore shape would also have occurred in the low density wheat crop, where root number and spread were much greater than in the high density wheat treatment.

Number of Pores

It was hypothesised that lodging would compress pore space thereby reducing the number of pores, or it would have the effect of creating more, smaller sized pores. There was no significant effect of lodging, however the interaction for soil type and lodging had a significant effect on reducing the number of pores in all three soils (p <0.05), with soil type having the most significant effect (p<0.001). The greatest differences between lodged and unlodged samples were in the high density wheat treatments in the clay soil, where pore number was reduced by approximately 50%. The smallest reduction of pores at high density occurred in the sandy loam soil, where lodging reduced pore numbers by approximately 3%. For the barley plots, at high density, lodging reduced the number of pores in the clay and silty loam, but the greatest effect of lodging occurred in the sandy loam where lodging increased pore number by approximately 25%. At low density, the greatest reduction in pore number occurred in the clay and sandy loam, with minimal pore reduction in the silty loam.

The results for total number of pores were variable across the three soil types with the greatest reduction of 50% in the high density wheat treatments of the clay soil. Given that clay soils are commonly composed predominantly of micropores (not measured here) it is perhaps not surprising that pore reduction was greatest in this soil. In a similar vein, the lowest pore number reduction in the sandy loam was also expected, due to the majority of pores in this coarse textured soil being much larger and thus fewer pores to be reduced. Crop density had variable effects on the effect of lodging on the number of pores. Generally at high density, the lodged treatments had a reduction in pore number with the exception of wheat in the silty loam and barley in the sandy loam. Possible explanations for this could be related to the root density of the crops, with roots occupying more pore space at high density compared to low density crops, meaning that there are fewer pores around the root mass in high density crops.

Figure 5: Mean pore area for each of the soil and crop types for lodged (L)/unlodged (UL) treatments.

The interaction between soil type and lodging had a most significant (p<0.05) effect on the mean pore area of the soils, but there was no significant effect for lodging alone. As shown in Figure 5, for the high density treatments lodging has the greatest effect on pore area in the clay, where the mean pore area was reduced from 2.5 mm² to 1 mm². In the silty loam, mean pore area was increased by 1 mm² compared to the unlodged sample, with least effect on the sandy loam where pore area remained constant at 0.42 mm² – 1.0 mm² in all treatments and soil types. In the barley treatments, at high density in both the clay and silty loam, lodging increased mean pore area compared to the unlodged samples. Lodging had the greatest effect in the clay, where pore area was increased from 0.91 mm² to 1.27 mm². In both the sandy and silty loam soils the pore area appears to have been little affected by lodging as both soils record a mean pore area of 0.5 mm²and 2 mm² respectively.

It was hypothesised that lodging would reduce the mean pore area in each of the soils through compaction of pores. It appears from the results that the greatest effect of lodging on reducing pore area occurred in the high density wheat treatment for the clay; this was not surprising given that pore shape became more elongated with lodging for the same soil type and plant treatment. However, in the remaining treatments for the clay soil, lodging increased the mean pore area, indicating that the lodging process increased the pore sizes. However, this was not reflected by pore shape where there was no significant change in the mean pore shape for the remaining treatments in the clay. The mean pore area results for the sandy and silty loam soils indicate that lodging does not significantly affect the pore areas.

Conclusions

Quantified evidence from the undisturbed samples of lodged and unlodged plants grown in clay, sandy and silt loam soils illustrated that lodging may have a significant effect on soil morphological changes. The impacts of lodging were greatest in the clay soil with regards to all measurements but significant changes were also revealed in the sandy loam. It has also been revealed that lodging significantly altered the number of pores across all treatments irrespective of soil type, whilst lodging had the most significant impact on pore shape under high density wheat treatments in the clay soil, where pore shape was changed from rounded to irregular. Mean pore area results showed that the least structural disturbance from lodging occurred in the sandy loam soil, where results were generally less variable than in the clay and silty loam soils. It can be concluded that lodging did alter soil morphology but greater research into the lodging process of particular exact areas of the root-soil complex is required to fully establish the nature of the relationships between soil type, crop density, cereal species and the type of lodging that occurs. Further work will analyse the soil within 2 cm x 2 cm sections in order to gain an understanding of the spatial variation in the lodging induced changes that occur in each of the three soils.

References

Baker CJ, Berry PM, Spink JH, Sylvester-Bradley R, Clare RW, Scott RK, Griffin JM, (1998) A method for the assessment of the risk of wheat lodging. Journal of Theoretical Biology 194, 587-603.

Berry PM, Griffin JM, Sylvester-Bradley R, Scott RK, Spink JH, Baker CJ, Clare RW, (2000) Controlling plant form husbandry to minimise lodging in wheat. Field Crop Research. 67, 59-81.

Berry PM, Sterling M, Baker CJ, Spink JH, Sparkes DL, (2003) A calibrated model of wheat lodging compared with field measurements. Agriculture and Forest Meteorology 119, 167-180.

Berry PM, Spink JH, Sylvester-Bradley R, Pickett A, Sterling M, Baker C, Cameron N, (2002) Lodging control through variety choice and management. In: Proceedings of the 8^th Home-Grown Cereals Association R&D Conference on Cereals and Oilseeds. HGCA, London, pp. 7.1-7.12.

Berry PM, Sterling M, Spink JH, Baker CJ, Sylvester-Bradley R, Mooney SJ, Tams AR, Ennos AR, (2004) Understanding and reducing lodging in cereals. Advances in Agronomy 84, 217-271.

Crook MJ, Ennos, AR, (1993) The mechanics of root lodging in winter wheat (Triticum aestivum L.). Journal of Experimental Botany 44, 1219-1224.

Easson DL, Pickles SJ, White, EM, (1993) A study of lodging in cereals. HGCA Report No. 52. HGCA, London.

Ennos AR, (1991) The mechanics of anchorage in wheat (Triticum aestivum L.) II. Anchorage of mature wheat against lodging. Journal of Experimental Botany 42, 1607-1613.

Graham J, (1983) Crop lodging in British wheats and barleys. Ph.D. Thesis, University of Reading, UK.

Mooney SJ, Holden NM, Ward SM, Collins JF, (2000) The use of quantified structural descriptors to physically characterise stockpiled milled peat of different genetic origins. Geoderma 95, 15-32.

Mulder EG, (1954) Effect of mineral nutrition on lodging in cereals. Plant and Soil 5, 246-306.

Neenan M, Spencer-Smith JL, (1975) An analysis of the problem of lodging with particular reference to wheat and barley. Journal of Agricultural Science Cambridge 85, 495-507.

Pinthus MJ, (1973) Lodging in wheat, barley and oats: the phenomenon, its causes and preventive measure. Advances in Agronomy 25, 209-263.

Sterling M, Baker CJ, Berry PM, Wade A, (2003) An experimental investigation of the lodging of wheat. Agriculture and Forest Meteorology 119, 149-165.

Tams AR, Mooney SJ, Berry PM, (2004) A method for rapidly impregnating and preserving soil structure in the field to enable recovery of ‘undisturbed’ samples (Submitted to Geoderma).

Thomas WD, (1982) Plant Growth Regulators. Yield of Cereals Course Papers, 1982. NAC Cereal Unit, RASE, Stoneleigh, UK, pp. 78-95.

White EM, McGrael ASL, Ruddle O, (2003) The influence of variety, year, disease control and plant growth regulator application on crop damage, yield and quality of winter oats (Avena sativa). Journal of Agricultural Science Cambridge 140, 31-42.