Archived at

Dijon - 31 août - 2 septembre 2004





Effect of repeated hoeing on growth of Cirsium arvense


E. Graglia, B. Melander, H.Grøndal & R. K. Jensen


Danish Institute of Agricultural Sciences, Department of Crop Protection

Research Centre Flakkebjerg, 4200 Slagelse, Denmark



Summary: Infestation with Cirsium arvense in organic cropping is an increasing problem in many parts of Europe. In the present study, repeated hoeing during the first part of the growing season is suggested as a mechanical control strategy, aiming at diminishing the regenerative capacity of C. arvense. Results showed a linear relationship between the number of hoeing passes and the aboveground biomass of C. arvense in the subsequent year. However, the strong effect on the biomass of C. arvense was not followed by an increased crop yield. A low initial density of C. arvense, which did not cause any measurable yield loss, most likely explains this.



Effets de binages répétés sur la croissance du chardon des champs (Circium arvense)

Résumé : L’infestation des cultures bilogiques par le chardon des champs est un problème croissant en Europe. Dans cette étude, une stratégie de binages répétés durant la première partie de la période de croissance est proposée pour diminuer le potentiel de régénération du C. arvense. Une relation linéaire a été observée entre le nombre de passages de bineuse et la biomasse aérienne du chardon au cours de la campagne suivante. Cependant, l’effet important sur la biomasse de chardon n’a pas induit d’effet sur le rendement de la culture. Ce résultat est probablement lié à un faible niveau initial d’infestation par le chardon sur la parcelle d’essai, sans effet mesurable sur le rendement.







Cirsium arvense L. (Scop) is a troublesome weed wherever it is found in agricultural ecosystems (Donald, 1994; Skinner et al., 2000). In conventional agricultural systems, targeted use of herbicide during the growing season can provide satisfactory control, whereas the problem tends to increase in organical growing systems. New plants can be recruited from seeds as well as small root fragments, and just one new plant has the potential to infest large areas within two to three years after establishment (Nuzzo 1997). This is due to the very expansive root system and the ability to form new aerial shoots from root buds. Furthermore, the high presence of labile carbohydrates in the roots of C. arvense enables the plant to regenerate even from root fragments 10 mm long (Hamdoun, 1972). Hence, disruption of roots from a single mechanical weed control event could facilitate an increased infestation (Bostrom & Fogelfors, 1999). It has been suggested that the amount of labile carbohydrates varies across the season (McAllister & Haderlie, 1985) and further, that the minimum dry weight of underground regenerative organs, the time when the plant is most susceptible to removal of aboveground plant tissue, is when the aerial shoot has approximately eight expanded leaves (Gustavsson, 1997). In this paper it is hypothesized that the use of repeated mechanical control events, removing aboveground biomass, are likely to deplete the carbohydrates of the root system, with a subsequently decreased regrowth capacity, measured as a decreased aboveground biomass the following year.

Two field experiments were carried out to investigate how the use of repeated hoeing during the first part of the growing season one year influences the growth of C. arvense and the yield of spring barley the year after the treatment.



Materials and Methods


Experimental site and set-up


The experiments; A (2001-2002) and B (2002-2003), were conducted on a sandy loam near the Research Centre Flakkebjerg and consisted of four blocks (replicates) with a gross plot size of 12.5 m x 2.5 and a net plot size of 10 m x 1.5 m. The field had been organically grown for more than 10 years and had a well-established population of C. arvense. The experimental set-up consisted of three levels of hoeing and an untreated control, making up a total of 16 plots in each experiment.


Experimental procedures


Spring barley and a catch crop consisting of white clover were sown at a row distance of 24 cm on 16 April 2001 in experiment A and on 8 April 2002 in experiment B. Just after sowing pelleted chicken manure was applied corresponding approximately to 70 kg N/ha. Hoeing, using a “ducks foot” blade 10 cm wide, was carried out whenever C. arvense had reached a height of 10 cm; in Exp. A until 21 May, 19 June and 19 July respectively in the three treatments and in Exp. B. until 22 May, 17 June and 17 July, respectively. The maximum height of C. arvense at the time of hoeing ensured a proper control by the hoe. For both experiments the chosen end dates gave rise to one, three and five hoeing events, respectively. After harvest the field was left untouched until November, where after the field was ploughed. The following spring i.e. the second experimental year, the plots were re-established from fix points along the field. All plots were fertilised with pelleted chicken manure, corresponding approximately to 70 kg N/ha, where after spring barley was sown. In the second experimental year no control of C. arvense was carried out.


Data collection 


Just before harvest in the second experimental year the height of all above ground shoots of C. arvense within the net plots was measured. From the gross plot, five above ground shoots of C. arvense from each height category of 0-10 cm, 11-20 cm and 21-30 cm etc. were harvested. The dry weight of these shoots was determined after drying at 85oC for 48 hours, and used as reference for estimation of above ground biomass of C. arvense within the net plots. Grain yield from the spring barley in the second experimental year was obtained by harvesting the net plots with a small-plot combine harvester. Grain dry matter was determined using a near infrared spectroscopy analyser (Foss Tecator, Infratec 1241) yield. Grain yield was adjusted to 85% dry matter.


Statistical analysis


To estimate the effect of repeated hoeing on the biomass of C. arvense a regression analysis was carried out using the SAS procedure PROC GLM (SAS, version 8.2). Comparison of regression lines between experiment A and B was carried out by including EXPERIMENT as class variable in the model statement of the PROG GLM procedure. Means of yield within experimental year were separated using Tukeys Test in the PROC GLM procedure. To ensure that no type II error was committed, when accepting our null hypothesis (Green, 1989), power analysis was performed according to Analyst Application (SAS, version 8.2).   





Fig. 1. Effect of repeated use of hoeing within one growing season on biomass of C. arvense the year after. Means ± se

Biomass of C. arvense


A linear relationship between the number of hoeing events and the biomass of C. arvense was found in both experiments, explaining 48% and 55% of the total experimental variation, respectively (Fig.1). The comparison of regression lines between experiments showed no difference between the slopes, but the intercepts were significantly different from each other. Any difference in the effect of the treatments is thus explained by the initial difference in the population size of C. arvense. 






Yield of spring barley


Table 1. Yield of spring barley, one year after the use of repeated hoeing. Means ± se. Within one column, means with the same letter are not significantly different.


                              Yield of spring barley (100 kg/ha)


Number of hoeing


Experiment A


Experiment B




29,4 ± 1,7a


34,2 ± 3,6a


29,3 ± 3,4a

36,9 ± 3,0a


33,9 ± 1,8a

39,8 ± 3,3a



29,0± 3,3a

40,1 ± 3,0a

The difference in hoeing intensity did not cause yield differences in spring barley grown the year after in any of the two experiments (Tab. 1). However, the power-test showed that the Power of the tests was below 0.8 for both years. Therefore the null hypothesis  - no difference between treatments - should be accepted cautiously.



Discussion and Conclusion


The results (Fig. 1) show that the amount of aboveground biomass of C. arvense the following growing season depends upon the number of hoeings. This supports our hypothesis that a continuous depletion of carbohydrates from the root system will diminish the regrowth capacity of the plant.

A seasonal pattern for labile root carbohydrates, with depletion during spring months as new shoots establish and replenishment through summer and fall has previously been reported (Welton et al., 1929; Otzen & Koridon, 1970). During springtime, the root system shifts from being a source to a sink for carbohydrates, and the dry weight of the root system will attain a minimum (Gustavsson, 1997) as will the regrowth capacity, which is the reason why treatments would be most effective at this time. However, the results from the present study are not sufficient for conclusions about the importance of the timing of the treatments. In contrast to the springtime campaign, Bourdôt et al. (1998) reported that late season mowing had the most severe impact on root biomass. However, the number of treatments confounded the effect of timing, as it was the case for the present study.

Even though C. arvense was the far most dominant weed species at the experimental site, a decreased aboveground biomass of the plant did not result in increased yield (Tab. 1), only a tendency to do so in experiment B, which had the largest population of C. arvense. Even though the statistical test showed that this increase was not significant, the power analysis at the same time made clear that an acceptance of the null-hypothesis could be erroneous. In another experiment adjacent to the present study, decreased aboveground biomass of C. arvense after repeated mowing increased the yield significantly (data not shown). However, the biomass of C. arvense in the control plots of that study was approximately 200 g/m2. It is therefore likely that the presence of C. arvense in the presented experiment caused a yield loss, but that the level was too low, to be revealed given the experimental set-up.

We have shown that repeated use of hoeing within just one growing season is likely to decrease the growth of C. arvense in the following season. However, it remains unknown whether or not the timing plays a role with respect to the effectiveness of the treatments. In addition, the time span in which an optimal weed control could be carried out would not only depend on the biology of the weed plant, but also of the chosen weeding tool. Not only from an academic point of view this would be interesting to look into. If the right timing, given different weeders, could diminish the resources, when trying to control C. arvense, this would be of great importance for practical use.






Bostrom U & Fogelfors H (1999) Type and time of autumn tillage with and without herbicides at reduced rates in southern Sweden: 2. Weed flora and diversity. Soil and Tillage Research 50, 283-293.

Bourdôt GW, Leathwick DM, Hurrell GA & Saville DJ  (1998) Relationship between areal shoot and root biomass in California thistle. In : Proceedings of the 51st New Zealand Plant Protection Conference, 28-32.


Donald WW (1994) The biology of Canada thistle (Cirsium arvense). Reviews of Weed Science, 6, 77-101.

Green RH (1989) Power analysis and practical strategies for environmental monitoring. Environmental Research, 50, 195-205.

Gustavsson AMD (1997) Growth and regenerative capacity of plants of Cirsium arvense. Weed Research 37, 229-236.

Hamdoun AM (1972) Regenerative capacity of root fragments of Cirsium arvense. Weed Research 12, 128-136.

McAllister RS & Haderlie LC (1985) Seasonal variations in Canada thistle (Cirsium arvense) root bud growth and root carbohydrate reserves. Weed Science 33, 44-49.

Nuzzo (1997) Element Stewardship abstract for Cirsium arvense. The Nature Conservancy, Arlington, Virginia, USA.

Otzen D & Koridon AH (1970) Seasonal fluctuations of organic food reserves in underground parts of Cirsium arvense (L) Scop and Tussilafo farfara L. Acta Botanica Neerlandica 19, 495.

SAS (1999). SAS System. SAS institute Inc., Cary, NC, USA.

Skinner K, Smith L & Rice P (2000) Using noxious weed lists to prioritize targets for developing weed management strategies. Weed Science 48, 640-644.

Welton FA, Morris VH & Hartzler AJ  (1929) Organic food reserves in relation to eradication of Canada thistles. Ohio Agricultural Experimental Station Bulletin. 441, 1-25.