Journal of Experimental Botany, Vol. 53, No. 367, pp. 333-340,
February 1, 2002
© 2002 Oxford University Press
Original Papers |
The role of root system architecture and root hairs in promoting anchorage against uprooting forces in Allium cepa and root mutants of Arabidopsis thaliana
Department of Biology, University of York, PO Box 373, York YO10 5YW, UK
Received 14 September 2001; Accepted 5 October 2001
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The role played by lateral roots and root hairs in promoting plant anchorage, and specifically resistance to vertical uprooting forces has been determined experimentally. Two species were studied, Allium cepa (onion) which has a particularly simple root system and two mutants of Arabidopsis thaliana, one without root hairs (rhd 2-1) and another with reduced lateral root branching (axr 4-2). Maximum strength of individual onion roots within a plant increased with plant age. In uprooting tests on onion seedlings, resistance to uprooting could be resolved into a series of events associated with the breakage of individual roots. Peak pulling resistance was explained in a regression model by a combination of a measure of plant size and the extent to which the uprooting resistance of individual roots was additive. This additive effect is termed root co-operation. A simple model is presented to demonstrate the role played by root co-operation in uprooting resistance. In similar uprooting tests on Arabidopsis thaliana, the mutant axr 4-2, with very restricted lateral development, showed a 14% reduction in peak pulling resistance when compared with the wild-type plants of similar shoot dry weight. The uprooting force trace of axr 4-2 was different to that of the wild type, and the main axis was a more significant contributor to anchorage than in the wild type. By contrast, the root hair-deficient mutant rhd 2-1 showed no difference in peak pulling resistance compared with the wild type, suggesting that root hairs do not normally play a role in uprooting resistance. The results show that lateral roots play an important role in anchorage, and that co-operation between roots may be the most significant factor.
Key words: Anchorage, lateral roots, root hairs, root system architecture, uprooting.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Anchorage failure is a problem in a wide range of crops, most notably cereals and timber trees, and can result in significant economic losses. When a plant is pulled vertically from soil, as in grazing, force is transmitted to the root system, which will fail at a point determined by the soil strength, the strength of the root–soil bond and the strength of the roots themselves in tension. In non-woody roots such failure generally occurs in proximal regions of the roots (Ennos, 1993
). However, the role of both lateral roots and root hairs in the anchorage of real roots remains unquantified, although Stokes et al. used wire models to predict that branching should increase uprooting resistance (Stokes et al., 1996). The rigid wire models are obviously not close mimics of non-woody roots, and indeed models that bent during uprooting behaved differently (Stokes et al., 1996). In most plants a single vertical force applied to a stem will be transmitted to numerous roots, either because of lateral branching or because of adventitious roots from the stem base. This allows more efficient transfer of the load to the soil because many narrow roots have a greater surface area than a single thick one (Ennos, 1993). Interactions among multiple roots may therefore further complicate the relationship between force applied and uprooting. The role of root hairs is even less clear; although they are often stated to play a role in anchorage, the fact that most roots break long before their tips are stressed suggests they do not (Ennos, 1993, 2000).
Quantifying the role of laterals and, more generally, of root architecture on anchorage will allow a clearer understanding of the relative importance of nutrient acquisition and anchorage in determining the evolution of the diversity of root system form (Fitter, 1999). It will provide the basic data from which to consider the potential for selection of crop plants resistant to uprooting.
Here three issues are addressed: (i) the role of multiple roots; (ii) the impact of laterals; and (iii) the contribution of root hairs to anchorage. Two species were chosen that have root systems suited to being used as model systems. Allium cepa L. (onion) cv. White Lisbon has a few thick, largely unbranched adventitious roots of uniform width that emerge from the base of the shoot. In sand culture, these roots do not produce root hairs. Experiments with A. cepa were therefore designed to measure uprooting forces in relation to the breakage of individual roots. Arabidopsis thaliana (L.) Heynh. (thale cress), was chosen because root mutants were available. Comparison of a root hair-deficient mutant with a wild-type A. thaliana allowed the role of root hairs in plant anchorage to be estimated. It was hypothesized that the mutant would have reduced uprooting resistance if root hairs played any role in anchorage. Finally, by comparing the anchorage of an Arabidopsis mutant with reduced lateral root production with a wild type, the effect of root architecture on uprooting resistance was measured.
It was hypothesized that reduced lateral production could have one of two contrasting impacts on anchorage: since it would result in deeper growth of the main root (Williamson et al., 2001), it would increase anchorage if lateral growth was unimportant, but reduce it if loss of the contribution of laterals outweighed this effect.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Features common to experiments using both species
Plants were grown from seed on a single batch of building aggregate (JH Walker Building Supplies, York) 80% sand, 20% small stones (>2 mm). Small amounts of silt and clay were present and pH was 7.7 in 1:2.5 0.01 M CaCl2. Pots were filled with building aggregate (referred to as ‘sand’) by hand and stones over about 1 cm in length were removed during pot filling. (There were very few stones between 5 mm and 1 cm in length.) A mesh square was placed in the bottom of the pot to stop loss of sand. Plants were grown in a growth room with fluorescent lights giving approximately 120 µmol m-2 s-1 PAR (photosynthetically active radiation) at plant height on a 16 h photoperiod 20/16 °C day/night. Plants were fed three times a week (details below) and additional deionized water was given as required. Prior to uprooting in a universal testing machine, pots were watered from above with tapwater until saturated and then allowed to drain for between 30 and 90 min. Except where specifically mentioned, the grip was a corrugated metal grip lined with thin pieces of rubber. If the plant shoot broke during testing, the data were discarded. In all cases dry weights refer to drying in an oven at 70–80 °C for 1 week.
Allium cepa
56 pots of 15 cm top diameter (volume approximately 1600 cm3) were filled with sand and randomly assigned to a harvest group (eight pots each, every week from 3–9 weeks after sowing). Due to space limitations not all harvest groups were started at the same time. Plants destined for harvest at age 5–9 weeks were started at the same time. After the early harvests (i.e. weeks 5 and 6), space was then available to start the plants destined to be harvested at 3 and 4 weeks after sowing. All plants were fed 40 cm3 half-strength Rorison nutrient solution (Booth et al., 1993) three times a week. One plant per week (two in week 6) was used for measurements of root strength. The remainder were tested for anchorage.
Allium cepa: individual root strengths
Plants were washed carefully out of soil and the individual adventitious roots originating from the stem base were removed for testing. The basal (i.e. nearest the shoot) 40 mm length was cut from each root. 10 mm at each end was glued (Loctite Superattak, a cyanoacrylate glue) between 0.1 mm thick steel plates. This left a 20 mm central section that was tested in a universal testing machine at a deformation rate of 20 mm min-1 imposing an initial strain rate of about 0.017 s-1. Roots under 40 mm long were ignored and if the root was over 80 mm long, a second 40 mm sample was taken from the basal end. During the preparation, the samples were allowed to dry out for about 30 min to allow the glue to adhere properly. Immediately prior to testing the roots were soaked in tap water for 5 min. This would have allowed complete recovery of turgor. Roots have to withstand drying occasionally, so it was considered that this uniform treatment was not markedly different from what would happen in life. Breaking force and the location of the break were recorded.
Allium cepa: vertical uprooting
Plants were pulled up at 500 mm min-1 in the Instron. Though a lower speed than would be produced by an uprooting herbivore, this was the highest cross-head speed that allowed faithful following of the load by the pen recorder. The speed was 25 times higher than was used in the single root strength experiments, but of course the difference in strain rate would be much less, because the roots were much shorter than the plant and its root system and, furthermore, much of the extension of the whole plant system would consist of straightening of roots, rather than extension of the tissue, which was what was measured on single roots. The clamps initially (weeks 5 and 6) were corrugated metal. This resulted in excessive damage to the shoot at the point of clamping, which was eliminated by lining the clamps with thin rubber. The number of broken main roots (excluding the much finer laterals) was counted using a hand lens. Once the number of broken roots was known, the sand was searched by hand until the entire root system had been recovered. Shoot and root dry weights were measured as above. From the force trace (an example of which is shown in Fig. 2), the peak pulling resistance PPR, the number of force drops greater than 0.15 N and the distance (on the x-axis in Fig. 2) between the first and second force drops over 0.15 N were measured. The distance between the force drop associated with the PPR and the next force drop over 0.15 N was also measured. (In Fig. 2, the peak pulling resistance occurs immediately before the first force drop, so these two measures are the same.)
|
Arabidopsis thaliana
Wild-type Columbia seed and the auxin-resistant mutant axr 4-2 (Hobbie and Estelle, 1995
) were kindly provided by Dr Ottoline Leyser, University of York. Rhd 2-1 (root hair-deficient; Schiefelbein and Somerville, 1990) gl1 (a glabrous leaf gene) in a Columbia background was provided by the Arabidopsis Biological Resource Centre (ABRC), Ohio State University. 120 pots, 9 cm top diameter were filled with sand and allotted randomly to the three seed types Columbia (wild type), axr 4-2 and rhd 2-1 gl1. Four seeds were sown per pot. Clear plastic sheet was laid over the pots for the first 6 d to stop the seedlings drying out. Feeding was three times a week with 20 cm3 one-fifth strength ATS feeding solution (Estelle and Somerville, 1987). Thirteen days after sowing, the seedlings were thinned to one per pot. Ten plants per seed type were harvested at 3, 4, 5, and 6 weeks after sowing. In half of them, shoot and root dry weight was measured. Because the line of demarcation between the root and shoot was not clear in some plants, the root was defined as including the hypocotyl. In the other half of the plants, shoot dry weight, peak pulling resistance and the size of the biggest single force drop on the force chart were measured. Plants were pulled out with a cross-head speed of 100 mm min-1 except in the week 6 harvest where it was 200 mm min-1. This difference in cross-head speed was an error. Fortunately but not surprisingly, see below, it seemed not to affect the results. When well-watered, the leaves of the basal rosette of A. thaliana lie flat against the sand. In order to get a good grip on the shoot, the leaves were bent upwards and the leaf stalks gripped. Throughout the experiment, rhd 2-1 gl1 plants were smaller than plants of the other genotypes and so the week six harvest of rhd 2-1 gl1 was split into two, at weeks 6 and 7, so as to obtain data for a similar plant size range as for the other genotypes. When PPR was compared for rhd2 gl1 plants in weeks 6 and 7, it was apparent that cross-head speed did not affect the data. |
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Allium cepa: individual root strength
Most roots (83/93) failed at or near a clamp. In 75% of the cases (63/83) this was at the lower clamp. This is a highly significant bias (
2=22.3, P<<0.001). Probably, roots tended to break near the bottom clamp because the roots became progressively weaker with distance from the stem, as observed earlier (Easson et al., 1995). However, there is no doubt that the load of some of the fractures near clamps would have been reduced by the local increase in stress induced by the clamps. The magnitude of this effect is likely to be similar for all the roots. The purpose of this particular experiment was merely to obtain an estimate of the strength of individual roots, and to determine whether this varied greatly with plants of different size. The results allowed the authors to choose a drop in load in the uprooting experiments that indicated that a root had broken (0.15 N, see next section), which was common to all ages of plants. The basal 40 mm sample was stronger than the next 40 mm sample (paired t-test, t39=5.27, P<0.001, mean difference 0.12 N). There was no relationship between the length of the root and the strength of the basal 40 mm. Although maximum strength of root samples (basal 40 mm) increased with shoot dry weight, the strength of the weakest root samples did not, resulting in a very large variance in the strengths of roots from large plants (Fig. 1).
|
Allium cepa: vertical uprooting
An example of an uprooting force trace (Fig. 2
) shows the characteristic form of the trace with a series of distinct force drops. There was a close relationship between the number of drops over 0.15 N in the force trace and the number of broken main roots, excluding the laterals (Fig. 3). The value of 0.15 N was chosen as the minimum size for a force drop a posteriori. A value of 0.1 N would allow inclusion of noise as drops and 0.2 N would exclude a significant fraction of roots breaking. Most (but not all) roots are stronger than 0.15 N in the basal 40 mm sample (Fig. 1). As the data on both axes in Fig. 3 can only take integer values, Kendall's non-parametric regression was applied (Sokal and Rohlf, 1995). The fitted regression line is the 1:1 line on Fig. 3. The relationships between log peak pulling resistance (PPR) and log shoot dry weight, log root dry weight and log plant dry weight are all similar (Table 1) and are all improved with the addition of a second predictor. The second predictor may be either of two closely related functions—either the distance (x-axis) between the first and second force drops or the distance between the drop associated with the peak force and the next force drop, both of which are negatively correlated with PPR. In 27 out of 32 cases, the peak force occurred at the beginning of the first drop, so that the two quantities are the same. In the subset of the data in which the number of broken roots equalled the number of force drops, the distance between the first and second force drops is the better predictor. In the entire data set the distance between the drop at the peak force and the next drop is the better predictor (Table 1). By excluding only one point from the full data set the distance between the first and second drops becomes again a better predictor than the distance between the force drop associated with the peak pulling resistance and the next force drop. The one excluded point had an exceptionally long distance between the first and second drops.
|
|
|
Arabidopsis thaliana
At all harvests, wild type-plants were larger than axr 4-2, which were larger than rhd 2-1 gl1. There was, however, no significant difference in the allometry of root and shoot weight (i.e. in a plot of ln root weight versus ln shoot weight) between the three Arabidopsis types nor between Columbia, rhd 2-1 gl1 (Fig. 4a
) or axr 4-2 (Fig. 4b) in the slope of relationship between log shoot dry weight and log peak pulling resistance.
|
However, axr 4-2 has a significantly lower (F1,37=11.2, P=0.002) intercept, than Columbia (Fig. 4b
). This is equivalent to axr 4-2 having a 14% lower peak pulling resistance than Columbia of the same shoot dry weight, at all plant sizes.
The ratio of the single largest force drop to the peak pulling resistance decreases with plant size (Fig. 5), but is greater in axr 4-2 than in Columbia at a given plant size. For analysis in a GLM, the square root of the y-axis (largest force drop divided by peak pulling resistance) was used to linearize the data. The regression of the square root of this ratio on log shoot dry weight, yields two parallel lines with different intercepts.
|
In plants with the same peak pulling resistance, the single biggest force drop is greater in axr 4-2 than in Columbia (Fig. 6
). Using GLM to compare the log largest force drop against log peak pulling resistance, parallel lines for axr 4-2 and Columbia are fitted, but with different intercepts.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
In Allium cepa (Fig. 3
) there was a 1:1 relationship between the number of drops in the force trace and the number of broken roots counted. When small numbers of roots broke (up to five), the number of drops was equal to or less than the number of broken roots observed. There are two reasons why there may be fewer drops than broken roots. One may be the breaking of weak roots (i.e. strength under 0.15 N). The other reason is that if two roots break at the same time, only one force drop will be recorded. With larger numbers of roots, the scatter about the 1:1 line is initially more even across both sides of the 1:1 line. This may be because as the plants get bigger and the total forces increase, the noise in the force trace increases. A constant force drop criterion (0.15 N) was kept because Fig. 1 shows no increase in minimum root strength with plant size. Drops in the force trace have been associated with root breakage by several authors (Blackwell et al., 1990; Ennos, 1991; Easson et al., 1995). Ennos showed a force trace of a wheat Triticum aestivum seedling being uprooted with three roots breaking and three drops in the force trace (Ennos, 1991).
The relationship of the distance trace between the first and second drops on the force trace and the peak pulling resistance (PPR) is explained by a simple model (Fig. 7). Essentially, the closer together the two drops are, the more the combined anchorage power of two roots can be regarded as additive; this additive property is referred to as root co-operation. If this were the whole explanation as to why PPR declines with the distance between the first and second drops, then the distance between the drop associated with the peak pulling resistance and the next drop should be a better predictor than the distance between the first and second drops. This is only partly true (Table 1). Of 32 plants, 27 had the peak pulling resistance at the first drop. A long distance between the first and second drops indicates that the first root to break is not contributing to peak pulling resistance (Fig. 8) and so peak pulling resistance is lower than one would expect for a plant of its size. However, when the distance becomes very long, small changes in this distance no longer bear any relation to peak pulling resistance (Fig. 8). This is the reason why, after eliminating one point with a long distance between first and second force drops, the better second predictor changes from the distance between the peak and next drops to the distance between the first and second drops (Table 1). Both of these variables measure root co-operation. The smaller the distance, the greater the co-operation. It is sometimes assumed that roots co-operate perfectly; for example, Ennos gives peak pulling resistance as sum of individual root strengths (Ennos, 1993). Coutts, however, noted that although the strength of the woody roots of trees in tension is not, in theory, subject to a reduction on subdivision, in practice it is reduced by unequal loading (i.e. one big root better than several small) (Coutts, 1983). Intuitively, co-operation between roots is only likely to be important when few roots are in tension. It is therefore possible that, in choosing a simple model in which there are few roots, an effect has been isolated that is important only in systems with few roots. However, trees may have only a few main laterals (Crook and Ennos, 1996) and as those in tension during lodging will be a fraction of the total, the number of main laterals in tension may be quite small. Root co-operation may therefore be important for trees too.
|
|
When A. thaliana was uprooted, PPR of Columbia was 14% greater than axr 4-2 for a given size of force drop, supporting the idea that laterals were contributing to anchorage, as already found in model systems (Stokes et al., 1996
). However, more roots broke in Columbia than in axr 4-2, hence the single biggest force drop is a greater proportion of the peak pulling resistance in axr 4-2 than in the wild type (Fig. 5). The strongest root in axr 4-2 (presumably the main axis) therefore appears to be stronger than the strongest root in the wild type when plants of the same peak pulling resistance are compared (Fig. 6). Stokes et al. showed that increased branching was associated with increased pullout resistance, but their experimental systems were very different from these described here as their models of roots were effectively infinitely strong (Stokes et al., 1995, 1996).
In Arabidopsis thaliana root hairs did not contribute to peak pulling resistance. Only a small proportion of the root pulled out with the shoot, implying that the value of peak pulling resistance in A. thaliana in this system is determined by root strength. Even in the case of rhd 2-1 gl1, the grip that the root is able to exert on the sand is greater than the strength of the root. Thus even if the root hairs in the wild-type Columbia do give better grip than in the root hair-deficient rhd 2-1 gl1, it does not result in greater peak pulling resistance in the wild type. The growing medium in this experiment was a loosely packed sand and as such likely to have a lower soil strength than most natural soils. It is unlikely therefore that root hairs contribute to anchorage in A. thaliana in most natural situations. Mutants of maize with much reduced root hairs have also been isolated (Wen and Schnable, 1994), but it is not known whether these mutants have anchorage similar to normal maize. Whilst A. thaliana roots are not strong enough to gain better anchorage from root hairs, the same may not be true for maize. Anchorage in the sense used above refers to whole plant anchorage against uprooting. Bengough and Mullins note that root hairs may anchor the root near its tip as the root tip forces its way through the soil (Bengough and Mullins, 1990). There is no contradiction in root tips being anchored by root hairs and at the same time for root hairs not to contribute to whole plant anchorage.
Two mutants of Arabidopsis are used to draw these conclusions. There is always a risk in using mutants that the genes in question may have pleiotropic effects. With axr 4-2 it is believed that the dominant phenotypic change is in root system architecture (Williamson et al., 2001), but it is possible that rhd 2-1 gl1, which grew more slowly than Columbia, may have differed in other relevant characteristics.
In summary, root co-operation has been demonstrated as an important variable in peak pulling resistance in Allium cepa. Lateral roots also contribute to anchorage and therefore interactions among laterals may also play a large role. However, it has not been possible to ascribe an anchorage function to root hairs. Understanding the processes or situations that promote root co-operation in anchorage might potentially enable manipulation of the plant or its environment to increase anchorage at no extra material cost to the plant.
|
Acknowledgments |
|---|
We are grateful to the Natural Environment Research Council for funding a studentship to the first author, and to Dr AR Ennos for comments on the manuscript; Dr Ottoline Leyser and the Arabidopsis Biological Resource Centre, Ohio State University kindly provided seed.
|
Notes |
|---|
1 To whom correspondence should be addressed. Fax: +44 (0) 1904 432 860. E-mail: ahf1{at}york.ac.uk
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Bengough AG, Mullins CE. 1990. Mechanical impedance to root growth: a review of experimental techniques and root growth responses. Journal of Soil Science 41, 341–358.
Blackwell PG, Rennolls K, Coutts MP. 1990. A root anchorage model for shallowly rooted sitka spruce. Forestry 63, 73–91.
Booth RE, Mackey JML, Rorison IH, Spencer RE, Gupta PL, Hunt R. 1993. ISP germination and rooting environments, sand, compost and solution culture. In: Hendry GAF, Grime JP, eds. Methods in comparative plant ecology: a laboratory manual. London: Chapman & Hall, 19–; 24.
Coutts MP. 1983. Root architecture and tree stability. Plant and Soil 71, 171–188.
Crook MJ, Ennos AR. 1996. The anchorage mechanics of deep-rooted larch, Larix europaeaxjaponica. Journal of Experimental Botany 47, 1509–1517.
Easson DL, Pickles SJ, White EM. 1995. A study of the tensile force required to pull wheat roots from the soil. Annals of Applied Biology 127, 363–373.
Ennos AR. 1991. The mechanics of anchorage in wheat Triticum aestivum. I. The anchorage of wheat seedlings. Journal of Experimental Botany 42, 1601–1606.
Ennos AR. 1993. The scaling of root anchorage. Journal of Theoretical Biology 161, 61–75.
Ennos AR. 2000. The mechanics of root anchorage. Advances in Botanical Research 33, 133–157.
Estelle MA, Somerville C. 1987. Auxin-resistant mutants of Arabidopsis thaliana with an altered morphology. Molecular and General Genetics 206, 200–206.
Fitter AH. 1999. Roots as dynamic systems: the developmental ecology of roots and root systems. In: Press M, ed. Plant physiological ecology. British Ecological Society Symposium No. 39, Oxford: Blackwell Scientific Publications, 115–131.
Hobbie L, Estelle M. 1995. The axr 4 auxin-resistant mutants of Arabidopsis thaliana define a gene important for root gravitropism and lateral root initiation. The Plant Journal 7, 211–220.[Web of Science][Medline]
Schiefelbein JW, Somerville C. 1990. Genetic control of root-hair development in Arabidopsis thaliana. The Plant Cell 2, 235–243.
Sokal RR, Rohlf FJ. 1995. Biometry, 3rd edn. New York: Freeman.
Stokes A, Fitter AH, Coutts MP. 1995. Responses of young trees to wind and shading: effects on root architecture. Journal of Experimental Botany 46, 1139–1146.
Stokes A, Fitter AH, Brain P, Coutts MP. 1996. An experimental investigation of the resistance of model root systems to uprooting. Annals of Botany 78, 415–421.
Wen TJ, Schnable PS. 1994. Analyses of mutants of three genes that influence root hair development in Zea mays (Gramineae) suggest that root hairs are dispensable. American Journal of Botany 81, 833–842.
Williamson L, Ribrioux S, Fitter AH, Leyser HMO. 2001. Phosphate availability regulates root system architecture in Arabidopsis thaliana. Plant Physiology 126, 875–882.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Stokes, A. Lucas, and L. Jouneau Plant biomechanical strategies in response to frequent disturbance: uprooting of Phyllostachys nidularia (Poaceae) growing on landslide-prone slopes in Sichuan, China Am. J. Botany, July 1, 2007; 94(7): 1129 - 1136. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. V. Minorsky Plant Physiology, June 1, 2002; 129(2): 438 - 439. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||