EFFECT OF LONG-TERM FALLOW TILLAGE METHOD ON WEED ABUNDANCE IN THE FIRST THREE YEARS OF FALLOW BEING BROUGHT INTO CULTIVATION

Ryszard Weber, Borys Hryńczuk; Bogdana Runowska-Hryńczuk

ABSTRACT

Over 2001-2003 there was researched an effect of various tillage methods on the reduction of weed infestation of a long-term fallow. The number of germinating weeds was analysed; the weeds had been obtained from a few-time-mixed soil samples collected from the top soil layers under the conditions of direct sowing, traditional ploughing, multistage ploughing and ploughing with a skim coulter. The results were compared with the abundance of weeds per 1 m² under field conditions. The log-linear analysis showed a considerable effect of weather conditions (years) on the abundance of germinating weeds in 0-5, 5-10, 10-15, 15-20, 20-25 cm layers. Multistage ploughing and ploughing with a skim coulter significantly decreased the number of germinating weeds in the top soil layers, as compared with the traditional plough and direct sowing. After two years of varied tillage methods considerably lower weed abundance was recorded for all the tillage variants, as compared with the results obta

Key words: fallow, weed infestation, tillage methods, log-linear analysis..

INTRODUCTION

The economic crisis in agriculture and ownership transformations over the last ten years in Poland have been responsible for fallows. On the territory of Western Europe, setting land aside is an element of rational reduction in overproduction combined with environment-friendly activities. In Poland, however, soils excluded from the agricultural utilization grow up with weeds and undergo a gradual degradation. High costs of traditional tillage as well as its negative impact on the soil environment had led to a more and more common application of simplified tillage over the last years. Direct sowing or conserving tillage decrease the rate of mineralization of organic matter and reduce considerably the wind and water erosion [34]. Simplified tillage methods, however, create much different conditions for the development of crops and weeds. Zero-tillage makes most weed seeds appear in the topsoil [35]. In direct sowing, only soil loosening of furrow with furrow openers creates unfavourable condi tions for the development of weeds [7,2]. The research showed that under zero-tillage, as compared with traditional ploughing, perennial weeds spread faster with a simultaneous reduction in the number of the abundance of annual species [6]. Depending on the stand, there was also noted an increased abundance of common lambsquarter (Chenopodium album), common chickweed (Stellaria media) and scentless mayweed (Tripleurospermum indorum) [23,35]. Various results of research which compared the effects of simplified tillage, traditional tillage or direct sowing are often a result of the use of the same herbicides in the tillage systems applied [7,11]. The effect of chemicals through a thick cover of mulch is more difficult, and a higher bioactivity of top soil layers in zero-tillage decreases the effectiveness of herbicides [30,36]. For that reason some reports show some greater weed infestation of fields under direct sowing than under ploughing [8,20,8]. When applying systemic herbicides , penetrating into the plant by overground parts, a much lower weed infestation was noted in zero-tillage as compared with that of the traditional method [21]. In simplified tillage or direct sowing herbicides are fixed by post-harvest residue or organic compounds of humus, which limits their effects considerably [3,13]. For that reason in direct sowing it is recommended to apply non-selective herbicides, e.g. based on glyphosate [9,26].

Crop rotation with the use of herbicides allows for avoiding a succession of weed species which have become resistant [25]. Traditional tillage and simplified tillage destroy emerging weeds to a large extent, however, at the same time enhancing seed germination due to a higher oxygen supply to soil and soil warming [1]. Weed infestation in zero-tillage can be reduced by the type of direct-sowing furrow opener. Disk opener limits the weed development more considerably than the tooth-type furrow opener, forming a narrower furrow. The literature available offers relatively little data on the distribution of weed diasporas in the soil profile, especially on long-term fallows [10,16]. Depending on the tillage method, the depth of weed diaspores distribution changes, as well as the effect of such factors as scarification and light reaction conditioning their emergence [17]. Taking off the top soil layer and placing it at specific depth, especially in the cultivation of long-term fallow, should li mit the potential weed infestation expansion. The aim of the present paper is to compare the effect of various tillage methods on the number of weed seeds and reduction of long-term fallow weed infestation.

MATERIAL AND METHODS

The research was carried out from 2000 at Jelcz-Laskowice on 12-year fallow, without succession towards shrubberies, grown up with perennials and annual plants. It was lessive soil of acid reaction – pH KCl – 4.6, formed from heavy loamy sand containing 0.95% C org. total. Following the fertilisation with lime (4 t of magnesium quicklime 65%), the total area allocated to the experiment was cross-harrowed using a heavy disk harrow. The soil richness in P and K following Egner-Riehm was average and amounted to 48 and 141 mgˇkg^-1, respectively. However, the content of Mg in soil was low and it amounted to 40 mgˇkg^-1. In mid September 2000, with a strong regrowth of weeds, the field was sprayed with a solution of glyphosate in a form of ammonium salt (2720 g ˇ 300 dcm^-1).

Bringing fallow under cultivation was initiated with four tillage methods (Table 1). In the spring of 2001 ‘Bahia’ maize was sown, followed by ‘Hetman’ oats and in the third year by ‘Pronto’ winter triticale. The experiment was set up using the randomized complete block design. Each object involved four replications. The plot was 150 m² in size.

The number of seeds in soil for respective experimental objects was determined annually after harvest. The samples were randomly taken to define the soil bulk density with cylinders 5 cm in diameter and 100 cm³ in volume. From each plot 0-5, 5-10, 10-15, 15-20, 20-25 cm deep four cylinders with undisturbed soil layer were sampled. The cylinders soil was dried at 20°C to obtain air-dry matter, and then moisten up to 60% of the maximum water capacity. The soil samples were placed in containers 8 cm in diameter under glass at 20°C to obtain constant moisture. The number over emergence was noted successively, and after two weeks soil samples were dried up again, mixed and moistened, starting the next cycle of observations. Over winter, over three weeks the soil samples temperature was periodically lowered (every 4 days) down to -5°C and then increased up to 5°C. The weed emergence observations for successive samples had been carried out up to June 30 of the following year. The results of calculations for four cylinder areas from each plot and from each depth were calculated per 1 m². Mean abundance of germinating weeds served as the basis of calculations.

The present paper pays a special attention to the interaction of respective factors analysed over the period researched, treating the number of weeds as a quality variable. To define the relationship between the abundance of germinating weeds in respective soil layers, weather conditions (2001-2003) and the tillage method, log-linear analysis was carried out, following Goodman and Everit [14,15]. The verification of the hypothesis on the factors studied being independent was evaluated using the Chi-square test. All the significant abundance deviations between that observed and expected in this analysis show an existence of a relation among the variables researched. In the log-linear model by the logarithmic transformation of the expected values, the model examined can assume a linear form in the simplest case using the following formula:

Ln(E_ij) = M. + λ_i^X + λ_j^Y + λ_ij^XY

where:

E_ij – expected values,
M. – general average based on an equal abundance of each variable,
λ_i^X – effect of i-value of variable X,
λ_j^Y – effect j-value of variable Y,
λ_ij^XY – effect of interaction of i-value of variable X and j-value of variable Y.

Although the model presented is very similar to the one of quantitative variables used in variance analysis, however the results interpretation method is different. The log-linear model allows a verification of the zero hypothesis which assumes no interaction between two or more variables analysed. It also makes it possible to, having rejected non-significant interactions, evaluate the effect of respective factors on the variability of the population studied. The calculations were made using STATISTICA 6.

RESULTS AND DISCUSSION

Mild winters over the research period analysed did not decrease the density of perennial weed species. The year 2000 recorded lower rainfall in April, June and July, as compared with the multi-year means (Fig. 1). High rainfall deficits and a high air temperature over spring and summer months of the year analyzed did not reduce the weed infestation of the 12-year fallow. In 2001 both rainfall over maize vegetation and the air temperature were higher than the multi-year means (Figs 1 and 2). Higher rainfall and high temperature were favourable for a considerable weed infestation. The year 2002 showed rainfall deficit in March, June and July. High temperature and low rainfall, as compared with the multi-year means over oat vegetation, were not favourable to a considerable weed infestation in the experiment. Conditions especially unfavourable to weed development were recorded in the year 2003. In each month analysed there were recorde d a considerably lower rainfall and higher daily air temperature, as compared with the previous years.

A phytosociological analysis of the fallow prior to tillage showed field weed infestation mainly with perennial weed species: wild tancy (Tanacetum vulgare (L.) Pal. Beauv.), mugwort (Artemisia vulgaris L.) and rhizomatous weeds: common couch-grass-quackgrass (Agropyron repens (L.) Pal. Beauv.), creeping thistle (Cirsum arvense (L.) Scop.), red sorrel (Rumex acetosella L.) (Table 2). The intensity of occurrence of other weed species was lower. In the first research year there was recorded a significantly lower number of weeds per 1 m² when applying multistage ploughing and ploughing with a skim coulter than in the other tillage variants (Table 3). The species which prevailed included barnyardgrass (Echinochloa crus-galli (L.) Pal. Beauv.), quackgrass (Agropyron repens), field horsetail (Equisetum arvense L.) and creeping thistle (Cirsum arvense), especially under direct sowi ng and traditional tillage. Also in the second year of cultivation the number of weeds per 1 m² was considerably lower in the variant after multistage ploughing (Table 4).

However objects of direct sowing recorded significantly higher weed abundance. The weather in 2002 was favourable to a considerable weed infestation with hairy tare (Vicia hirsuta (L.) Gray), field pansy (Viola arvensis Murr.), and wild buckwheat (Polygonum convolvulus L.).

In the direct sowing variant perennial species and grasses were more abundant than in case of the other tillage methods. In the third year a much lower weed infestation per 1 m² was recorded for all the tillage methods than in the preceding years (Table 5). Multistage ploughing and ploughing with a skim coulter significantly lowered the abundance of weeds, as compared with the traditional tillage and direct sowing. The dominant species, especially in direct sowing, were: quackgrass (Agropyron repens), field horsetail (Equisetum arvense), mugwort (Artemisia vulgaris), and creeping thistle (Cirsum arvense).

In order to define the optimum model to test differences in the number of weed seeds for respective tillage variants there was calculated the Chi-square test value for the main effects without interaction. Then extended models were analysed, factoring in the two-factor and three-factor interactions (Table 6). Statistics calculated for the two- and three-factor model were significant (p < 0.01) and so one must reject the hypothesis stating that the number of germinating weeds does not depend on the years, tillage methods and soil sampling depth. One shall state at the same time that including the third order interaction into the model considered makes it fit more.

Table 7 shows which interactions analysed were significant in the model researched. The partial association informs if the interaction shows an effect when all the other effects of the same degree have already been included in the model. However marginal associations define the comparison of the model without any interactions with the model considering only a given interaction. The interaction shown between the tillage method and the soil sampling depth show a varied abundance of germinating weeds depending on the tillage variant and measurement depth. Also significant interactions with the tillage methods or measurement depths confirm a considerable effect of weather conditions on the number of germinating weeds. In order to fit the model to the abundance of germinating weeds iteration procedure was carried out.

The procedure was interrupted as the difference between the fitted and the observed marginal distributions was not greater than the convergence criterion = 0.01. In practice as a result of the above procedure most frequently the model excludes non-significant interactions between the variables researched.

The log-linear analysis forms also the so-called marginal tables of the current model fitted to the calculated number of germinating weeds in respective experimental objects. Model fitting is related to a calculation of expected values which reflect the marginal tables abundance. Analysing the expected values in Tables 8, 9 and 10, there was observed a lower total abundance of germinating weeds in the third year of cultivation, as compared with those of 2001 and 2002. Multistage ploughing limited the abundance of inferior plant species more considerably than the other tillage variants. Under direct sowing in the 0-5 cm layer, there were recorded higher abundances of weeds in the second and third year, as compared with traditional ploughing, multistage ploughing or ploughing with a skim coulter. However the traditional tillage method (ploughing) over the three years analysed was decreasing the number of germinating weeds in t opsoil.

The results presented show that in case of long-term fallow, both an intensive tillage and direct sowing after two years considerably limit the abundance and weed species diversity. The reports of other authors confirm that over three years under the conditions of direct sowing one can reduce considerably the seeds in topsoil [22,35]. Zero-tillage is much disturbed by perennial and monocotyledonous weeds whose control, also in other experiments, was very troublesome [4,11]. No mixing of soil in direct-sowing variant decreases the abundance of weeds despite a considerable number of seeds in topsoil, which is seen by considerable differences in the abundance of germinating weeds in the 0-5 cm layer in direct sowing and plough tillage, as compared with the results obtained per 1 m² under field conditions. Similarly Opic [22] in his research showed that decreasing the ploughing depth and repeated direct sowing result in an increased number of weed seeds in topsoil. Numerous reports b oth foreign [2,5,31] and domestic [12,24] show an increase in weed infestation of fields when exposed to direct sowing. However some research show that no mechanical treatments and a thick mulch layer limit the abundance of germinating weeds [32,33]. However shallow soil loosening in simplified tillage triggers weed germination [27,28,29]. Uneven supply of water to weeds under conserving tillage, due to seed shedding in the straw mulch layer, makes germinating weeds when exposed to a short rainfall deficit dry up [7,19]. Some weed species depend on the so-called reaction to light which triggers emergence in plants [16]. When exposed to no light, the weed seeds, although persisting in topsoil, remain in the state of dormancy not infesting crop plantations [7,19]. Limiting the weed infestation in case of simplified tillage plantations is also related to the mulch layer secreting substances stimulating a development of some weed species [30,31]. The above reports account for the fact of a considerable weed infes tation of objects in zero-tillage, as compared with direct sowing and traditional tillage. A higher abundance of some weed species in traditional tillage show a varied capacity of the plants studied to adapt to extreme environmental conditions. The above result coincides with the reports of other authors [2]. Reports of the last few years also reports on considerable differences in the competitiveness of cereal cultivars towards the existing weed populations [17]. With that in mind in the future a selection of the existing Polish cultivars should be made which would include that trait. The results given show that reclamation multistage ploughing most considerably limited the weed infestation in the fallow researched. However its cost makes it possible for the treatment to lower the production profitability as compared with the other tillage methods.

CONCLUSIONS

1. Reclamation multistage ploughing in the first year of bringing fallow into cultivation limited weed infestation more considerably than the other tillage methods.

2. Direct sowing after three years considerably lowered the weed abundance. However the intensity of occurrence of monocotyledonous and perennial species was higher than in the other tillage variants.

3. The share of respective weed species after varied tillage methods applied depended considerably on weather conditions and the crop over the research years analysed.

4. No mixing of soil under direct-sowing conditions decreases the abundance of weeds despite a considerable number of seeds in topsoil.

1. Amann A., 1992. Einfluss vin Saattermin und Grundbodenbearbeitung auf die Verunkrautung in verschiedenen Kulturen [Effect of sowing date and tillage on weed infestation of different crops.]. Dissertation, Universität Hohenheim [in German].

2. Arshad M.A., Gill K.S., Coy G.R., 1994. Wheat yield and weed population as influenced by three tillage systems on a clay soil in temperate continental climate. Soil Till. Res. 28, 227-238.

3. Beulke S., Malkomes H.P., 1996. Abbau der Herbizide Ethofumesat und Metazachlor in Böden mit unterschiedlichen C_org-Gehalten [Decreasing doses of Ethofumesat and Metazachlor herbicides in soils with a varied content of C _org] Z. für Pflanzenkrankheiten und Pflanzenschutz 15, 609-618 [in German].

4. Buhler D.D., Oplinger E.S., 1990. Influence of tillage systems on annual weed densities and control in solid-seeded soybeans. Weed Sci. 38, 158-165.

5. Braim M.A., Chaney K., Jodgson D.R., 1992. Effects of simplified cultivation on the growth and yield of spring barley on a sandy loam soil. Shoot growth and grain yield. Soil Till. Res. 22, 159-171.

6. Bräutigam V., 1993. Einfluss verschiedener Bodenbearbeitungssysteme auf Halmbasis-krankheiten des Getreides, die Unkrautentwicklung und – bekämpfung [Effect of varied tillage systems on the occurrence of cereal culm base diseases, weed development and control.]. Dissertation, Giessen [in German].

7. Cardina J., Regnier E., Harrison K., 1991. Long-term tillage effects of seed banks in three Ohio soils. Weed Sci. 39, 186-194.

8. Carter M.R., Sanderson J.B., Ivany J.A., White R.P., 2002. Influence of rotation and tillage on forage maize productivity, weed species, and soil quality of a fine sandy loam in the cool-humid climate of Atlantic Canada. Soil Till. Res. 67, 85-98.

9. Cremer J., Beestermöller H., Schutzler K., 1996. Zum derzeitigen Stand der Feldversuche mit glufosinat-ammonium in glufosinat – verträglichen Kulturen [Current status of field experiments with gluphosinatammonium in gluphosiant in resistant crop species.]. Z. für Pflanzenkrankheiten und Pflanzenschutz 15, 173-179 [in German].

10. Duer I., 1998. Zachwaszczenie w życie ozimym po jednorocznym ugorowaniu [Weed infestation in winter rye after a single-year fallow]. Bibl. Fragm. Agron. 5, 137-145 [in Polish].

11. Dycker J., Rux S., Knechtges H., 1992. Direktsat in semiariden Klimaten [Direct sowing in semi-dry climateang.]. Landtechnik 47 (7/8), 323-327 [in German].

12. Dzienia S., Piskier T., Wereszczaka J., 1998. Wpływ systemów uprawy roli na plonowanie i zachwaszczenie pszenicy ozimej [Effect of tillage systems on winter wheat yielding and weed infestation]. Rocz. Nauk Roln. 1-2, 37-42 [in Polish].

13. Düring R.A., Hummel H.E., 1994. Der Eibfluss differenzierter Bodenbearbeitung auf das Verhalten ausgewählter Herbizide im Boden [In:] Tebrügge F., Dreier M., Beurteilung von Bodenbearbeitungssystemen hinsichtlich ihrer Arbeitseffekte und deren langfristigen Auswirkungen auf den Boden [Effect of a varied tillage on the activity of some herbicides in soil.]. Giessen, 65-82 [in German].

14. Everitt B.,S. ,1977. The analysis of contingency tabels. Chapman & Hall, London.

15. Goodman L.A., 1978. Analyzing Qualitative/Categorical Data : Log-Linear Models and Latent-Structure Analysis, ed. by Jay Magidson Cambridge.

16. Hochół T., Łabza T., Stupnicka-Rodzynkiewicz E., 1998. Zachwaszczenie wieloletnich odłogów w porównaniu do stanu na polach uprawnych [Weed infestation of long-term fallows, as compared with the condition in fields]. Bibl. Fragm. Agron. 5, 115-125 [in Polish].

17. Juroszek P., Gerhards R., Kűhbauch W., 2002. Photobiologische Unkrautregulierung annueller Ackerunkräuter [Photobiological (light) control of weed infestation of annual field weeds]. J. Agronom. Crop Sci. 188, 389-397 [in German].

18. Klümper H., Gerhards R., Kühbauch W., 1996. Einfluss des Lichtes auf die Keimung von Unkrauten [Effect of light on weed germination]. Z. für Pflanzenkrankheiten und Pflanzenschutz 15, 71-73 [in German].

19. Lemerle D., Smith A., Verbeek B., Rudd S., Martin P., 2000. Breeding for competitive cultivars of wheat. 3^rd International Weed Sci. Congress, IWSC FOZ DO IGUASSU, Brazil, 75.

20. Mills J.A., Witt W.W., 1989. Effect of tillage systems on the efficacy and phytotoxicity of imazaquin and mazethapyr in soybean (Glycine max. L). Weed Sci. 37, 233-238.

21. Mohler C.L., 1993. A model of the effects of tillage on emergence of weed seedlings. Ecological applications 3, 53-73.

22. Opic J., 1998. Wpływ głębokości orki i siewu bezpośredniego na liczbę nasion chwastów w glebie [Effect of ploughing depth and direct sowing on the number of weed seeds in soil]. Rocz. Nauk Roln. A 112 (1-2), 110-121 [in Polish].

23. Orzech K., Wanic M., Nowicki J., 1999. Wpływ zróżnicowanej uprawy roli na zachwaszczenie i plonowanie pszenicy ozimej w warunkach gleby średniej [Effect of varied tillage on winter wheat weed infestation and yielding in average-quality soil]. Fol. Univ. Agric. Stetin., Agricultura 74, 141-146 [in Polish].

24. Roszak W., Radecki A., Witkowski F., 1991. Badania nad możliwością zastosowania siewu bezpośredniego w warunkach Polski Centralnej [Research into the potential of direct sowing in Central Poland]. Rocz. Nauk Roln. A 109 (2), 143-156 [in Polish].

25. Rubin B., 1996. Herbicide-resistant weeds – the inevitable phenomenon; mechanism, distribution and significance. Z. für Pflanzenkrankheiten und Pflanzenschutz 15, 17-32 [in German].

26. Rupalla R., 1997. Kreuz und quer verflochten [Varied soil tillage]. DLG - Mitteilungen 1, 57-59.

27. Schmidt W., Doll D., Nitzsche O., 1999. Erfahrungen mit konservierender Bodenbearbeitung in Sachsen [Results and experiments in conserving tillage in Saxony]. Neue Landwirtschaft 5, 2-5 [in German].

28. Schmidt W., Eikenbusch J., Sieberhein K., 1998. Mit der richtigen Strategie bleiben die Kosten niedrig [Oriented tillage strategy aimed to reduce costs]. Top Agrar 8,1-3 [in German].

29. Schmidt W., Stahl H., 1999. Winterraps mit Erfolg ohne Pflug bestellen [Success of zero tillage in winter rape]. Bauern Zeitung – Sonderdruck Sächsische Landesanstalt für Landwirtschaft, 1-4.

30. Shumway D.L., Koide R.T., 1994. Seed preferences of Lumbricus terrestris L. App. Soil Ecology 1, 11-15.

31. Streit B., Stamp P., Richner W., 2000. No-tillage in cool and humid climate: weed populations and weed control. 3^rd International Weed Sci. Congress, IWSC FOZ DO IGUASSU, Brazil.

32. Teasdale J.R., Beste C.E., Potts W.E., 1991. Response of weeds to tillage and cover crop residue. Weed Sci. 39, 195-199.

33. Vidal R.A., Baumann T.T., 1994. Straw density in no-till affects soybean – weed interference. Proc. 3^rd ESA Congress, Padova (Italy), 268-269.

34. Weber R., 2002. Wpływ uprawy zachowawczej na ochronę środowiska [Effect of sustainable land cultivation on environmental protection]. Post. Nauk Roln. 1, 57-67 [in Polish].

35. Wrzesińska E., Dzienia S., Wereszczaka J., 2003. Wpływ systemów uprawy roli na ilość i rozmieszczenie nasion chwastów glebie [Effect of tillage systems on the abundance and distribution of weed seeds in soil]. Acta Sci. Pol., Agricultura 2(1), 169-175 [in Polish].

36. Zwerger P., 1994. Wie lange kannUnkrautsamen überleben? [How long can weed seeds survive in soil?]. Lohnunternehmen in Land – und Forstwirtschaft 49 (11), 544-545 [in German].

Responses to this article, comments are invited and should be submitted within three months of the publication of the article. If accepted for publication, they will be published in the chapter headed ‘Discussions’ in each series and hyperlinked to the article.