Preventive and cultural methods for weed management - Paolo Bàrberi

INTRODUCTION

In many agricultural systems around the world, competition from weeds is one of the major factors reducing crop yield and farmers’ income. In developed countries, despite the availability of high-tech solutions (e.g. selective herbicides and genetically-modified herbicide-resistant crops), the share of crop yield loss to weeds does not seem to reduce significantly over time (Cousens and Mortimer, 1995). In developing countries, herbicides are rarely accessible at a reasonable cost, hence farmers often need to rely on alternative methods for weed management.

Worldwide limited success in weed control is probably the result of an over-simplification in tackling the problem. Too much emphasis has been given to the development of weed control tactics (especially synthetic herbicides) as ‘the’ solution for any weed problems, while the importance of integrating different tactics (e.g. preventive, cultural, mechanical, and chemical methods) in a cropping system-based weed management strategy has long been neglected.

Integrated weed management is based on knowledge of the biological and ecological characteristics of weeds to understand how their presence can be modulated by cultural practices. Based on this knowledge, the farmer must first build up a global weed management strategy within her/his cash crop sequence, and then choose the best method of direct weed control during crop growing cycles. Besides this, it must be remembered that weed management is always strictly embedded in crop management itself. As such, the interactions between weed management and other cultural practices must be duly taken into account. For example, the inclusion of cover crops in a crop sequence is an interesting way to integrate weed management with nutrient management in low-external input systems, with additional benefits on other important agro-ecosystem properties (e.g. soil fertility, soil moisture retention, biodiversity, etc.).

HOW TO IMPLEMENT AN EFFECTIVE WEED MANAGEMENT STRATEGY

A long-term effective weed management strategy is based on the practical application of the ecological concept of ‘maximum diversification of disturbance’, which means diversifying crops and cultural practices in a given agro-ecosystem as much as possible. This results in a continuous disruption of weed ecological niches (Liebman and Davis, 2000) and hence in a minimized risk of weed flora evolution towards the presence of a limited number of highly competitive species. Besides this, a highly diversified cropping system also reduces risk of the development of herbicide-resistant weed populations.

In practice, weed management strategies should integrate indirect (preventive) methods with direct (cultural and curative) methods. The first category includes any method used before a crop is sown, while the second includes any methods applied during a crop growing cycle. Methods in both categories can influence either weed density (i.e. the number of individuals per unit area) and/or weed development (biomass production and soil cover). However, while indirect methods aim mainly to reduce the numbers of plants emerging in a crop, direct methods also aim to increase crop competitive ability against weeds.

Preventive methods include crop rotation, cover crops (when used as green manures or dead mulches), tillage systems, seed bed preparation, soil solarization, management of drainage and irrigation systems and of crop residues.

Cultural methods include crop sowing time and spatial arrangement, crop genotype choice, cover crops (when used as living mulches), intercropping, and crop fertilization.

Curative methods include any chemical, physical (e.g. mechanical and thermal) and biological methods used for direct weed control in an already established crop. A list of the main methods that can be used in an integrated weed management strategy is shown in Table 1.

Hereafter, the main effects on weeds of preventive and cultural methods are described, trying especially to highlight their possible interactions, which are not always easy to predict in the field. Curative methods are not treated here; however, it must be stressed that the effectiveness of any of them can be expected to increase if preventive and cultural methods are concurrently applied.

PREVENTIVE METHODS

Crop rotation

Differentiation of crops grown over time on the same field is a well-known primary means of preventive weed control. Different crops obviously bring about different cultural practices, which act as a factor in disrupting the growing cycle of weeds and, as such, preventing selection of the flora towards increased abundance of problem species (Karlen, 1994). In contrast, continuous cropping selects the weed flora by favouring those species that are more similar to the crop and tolerant to the direct weed control methods used (e.g. herbicides) via repeated application of the same cultural practices year after year.

In addition, continuous cropping can negatively interact with tillage systems and shift the weed flora towards a troublesome composition. For example, in continuous winter cereal cropping in temperate regions, minimum tillage can cause the dominance of grasses with low-dormant seeds, such as Alopecurus myosuroides and Bromus spp., to occur after a few years. (Froud-Williams, 1983). In these cases, the consequent higher use of graminicides acts as an additional selection factor for the weed flora and can also accelerate the selection of herbicide-resistant biotypes. To recover highly degraded floristic situations such as the one just pictured, it is imperative to rotate cereals with crops having a different growing period, as well as to plough the soil from time to time to disadvantage low-dormant grass species whose seeds are usually unable to emerge from deep soil layers. If there is a long fallow period between the cereal and the next crop, this can be exploited to cultivate the soil to stimulate the emergence of problem weeds, which are then destroyed by additional cultivation or by herbicides.

Rotation between crops having the same growing period, although certainly preferable to continuous cropping, is not as successful as rotation between crops with different growing cycles in reducing the number of weeds emerging in the field. Compared to weed-density reduction, the effect of crop rotation on weed biomass reduction is less systematic because it depends on factors such as the following:

• the competitive ability of crops included in the rotation;

• the effectiveness of direct weed-control methods (e.g. herbicides), and

• the frequency of tillage and cultivation.

Table 1. Classification of cultural practices potentially applicable in an integrated weed management system, based on their prevailing effect.

Cover crops (used as green manures or dead mulches)

Inclusion of cover crops in a rotation in the time frame between two cash crops is another good preventive method to be used in a weed management strategy. Cover crops do not give a marketable yield but, by extending the period in which the soil remains covered by vegetation, exert a series of beneficial effects on the agro-ecosystem, such as optimization of natural resource use (solar radiation, water, soil nutrients), reduced water runoff, nutrient leaching and soil erosion, and, last but not least, weed suppression (Lal et al. 1991).

Cover crop effects on weeds largely depend upon cover crop species and management, following cash crop, and weed community composition (Bàrberi and Mazzoncini, 2001). Weed suppression is exerted partly through resource competition (for light, nutrients and water) during the cover crop growing cycle, and partly through physical and chemical effects that occur when cover-crop residues are left on soil surface as a dead mulch or ploughed down and hence used as green manure (Mohler and Teasdale, 1993; Teasdale and Mohler, 1993). Interference with weeds, including competition, physical, and allelopathic effects, is generally higher when grasses or crucifers are used as cover crops than when legumes are used (Blum et al. 1997). Interference from cover crops and their residues is related to their occupation of ecological niches otherwise available for weeds. This is mostly a result of the sequestration of soil nutrients (especially N), to the release of allelochemicals (e.g. glucosynolates from crucifers and sorgoleone from Sorghum spp.) and to modifications of the soil microenvironment (Gallandt et al. 1999). Examples of highly weed suppressive cover crops are rye (Figure 1), sorghum, kale, rocket and mustard. In contrast, although direct weed suppression by legumes can be significant, their residual weed control effect is usually lower because the high quantity of N released from their residues after cover crop destruction stimulates weed emergence, especially when legumes are used as a green manure (Blum et al. 1997).

When cover crops are used as dead mulch (i.e. they are left to decompose on soil surface), weed suppression seems mostly to be the result of the physical effects of mulch, rather than to nutrient- or allelochemical-mediated effects (Teasdale and Mohler, 2000). In particular, weed suppression seems directly related to the Mulch Area Index (mulch area divided by soil unit area), which influences light extinction through the mulch and consequently weed seed germination. Small-seeded weed species appear to be more sensitive than large-seeded species to mulch physical effects as well as to allelochemicals (Liebman and Davis, 2000). Timely sowing of cover crops is very important to enhance biomass production and hence to increase their weed suppression potential.

Cover crops can also interact with other biota; for example, they promote the establishment of vesicular-arbuscular mychorrhizae, which in turn may shift weed flora composition by favouring mychorrhizal plant species at the detriment of non-mychorrhizal species (Jordan et al. 2000).

Figure 1. High weed suppression exerted by a rye cover crop

(Photo: P. Bàrberi).

TILLAGE SYSTEMS

The effect of primary tillage on weeds is mainly related to the type of implement used and to tillage depth. These factors considerably influence weed seed and propagule distribution over the soil profile and therefore they directly affect the number of weeds that can emerge in a field.

Mouldboard ploughing is very effective in reducing weed density and hence it is an important preventive method where farmers are forced (or are willing) to use partially suppressive direct weed control methods (e.g. mechanical weeding), and reduces the labour needed for subsequent hand-weeding. In contrast, with non-inversion tillage (especially with no tillage) weed seeds are only partially buried and therefore they are mainly distributed in the upper soil layer, from which they can easily germinate and give origin to established plants.

Theoretically, if direct weed control was effective enough to reduce weed seed shedding (S), non-inversion tillage systems should reduce weed density over time to a greater extent than plough-based systems. This should happen because of higher weed seed bank depletion (D) in non-inverted soil, driven by higher emergence rates and environmental conditions (related to lack of seed burial) not conducive to seed secondary dormancy; and by higher seed predation by soil fauna. In terms of weed population dynamics, a reduction in population size occurs if D > S, a situation that is very rarely encountered with non-inversion tillage because on-field weed control is rarely complete, therefore weeds have a very good chance of setting seeds and replenishing the soil seed bank. For this reason, weed densities in minimum- and no-tillage systems are invariably higher than in plough-based systems (Froud-Williams, 1988; Cardina et al. 1991; Spandl et al. 1999). Weed seed bank data taken in a long-term experiment in which four tillage systems were used for 12 consecutive years in a continuous winter wheat or a pigeon bean-winter wheat rotation showed that total weed seedling density was higher in no tillage, minimum tillage (i.e rotary harrowing at 15 cm depth), and chisel ploughing (at 45 cm depth) in the 0-15, 15-30, and 30-45 cm soil layers respectively (Bàrberi and Lo Cascio, 2001). Density in the whole (0-45 cm) layer did not significantly differ among tillage systems, but in no tillage more than 60 percent of total seedlings emerged from the surface layer, compared to an average 43 percent in the other tillage systems (Figure 2). Crop rotation did not influence either weed seed bank size or seedling distribution among soil layers, and had a small influence on major species abundance. The weed seed bank was dominated (> 66 percent of total density) by Conyza canadensis and Amaranthus retroflexus, which thrived with chisel ploughing and no tillage, respectively. Among other major species, Bilderdykia convolvulus and Chenopodium album were mainly associated with mouldboard ploughing, Papaver rhoeas and Portulaca oleracea with minimum tillage, and Lolium multiflorum and Veronica spp. with no tillage. Results suggest that, although substitution of mouldboard ploughing by non inversion tillage (especially by minimum tillage) may not result in increased weed problems in the long-term, use of no tillage is likely to increase weed infestations because of higher seedling recruitment from the topsoil, and consequently an increased requirement for herbicide application. Use of no tillage can be desirable in the tropics because these conditions would exacerbate weed-control problems.

Figure 2. Percent weed seedling distribution over soil layers in mouldboard ploughing at 45 cm depth (P 45), chisel ploughing at 45 cm depth (CP 45), rotary harrowing at 15 cm depth (RH 15), and no tillage (NT) after 12 consecutive years’ application of the different tillage systems (after Bàrberi and Lo Cascio, 2001, modified; data are means of two crop rotations). For each layer, bars labelled with the same letter are not significantly different at P £ 0.05 (LSD test).

Disturbance posed to weeds by tillage is dependent more on the type of implement than on tillage depth. Tools that do not invert the soil (e.g. chisels) increase weed density and shift weed flora composition towards an increased presence of biennials, perennials, and non-seasonal annuals. Most of these species are characterized by wind-dispersed seeds with reduced longevity and dormancy and are unable to emerge from deep soil layers (Zanin et al. 1997). Examples of species usually favoured by non inversion tillage or no tillage are Agropyron repens, Calystegia sepium, Lolium perenne and Plantago spp. (perennials), Digitaria sanguinalis, Lolium multiflorum, Setaria viridis and Thlaspi arvense (annuals).

Relative abundance of perennial species in a weed community is also favoured by reduced tillage frequency over a crop sequence. For example, the inclusion of a perennial ley in a crop rotation means that the soil is not tilled yearly. Lack of soil disturbance, coupled with higher control of annual weeds by repeated mowing in the ley, can shift weed community composition towards a higher presence of biennials and perennials. In contrast, plough-based systems seem to encourage some annual dicots such as Chenopodium album, Papaver rhoeas and Polygonum spp., although this effect is always modulated by the effectiveness of direct weed-control methods (e.g. herbicides) used in the crop rotation (Légère et al. 1993; Liebman et al. 1996).

In a given cropping system, weed density can be reduced to a greater extent when tillage methods change than when the same tillage system is used year after year. A long-term trial carried out at Pisa (Italy) in a soybean-winter wheat two-year rotation showed that by alternating mouldboard ploughing at 50 cm depth with minimum tillage (rotary harrowing at 15 cm depth) it was possible to reduce weed density in wheat compared with chisel ploughing (at 50 cm depth), minimum tillage or shallow ploughing (at 25 cm depth) when used every year (Bàrberi et al. 2001, Table 2). Use of minimum tillage for winter wheat and of deep ploughing for soybean was better than the opposite system because in the first case the weed community was mainly composed by less competitive species (Anagallis arvensis and Papaver rhoeas vs. Lolium spp., Polygonum aviculare and Veronica spp. in the second case). A very simple way to diversify the tillage system is to include in a rotation crops that require different tillage operations (e.g. cereals and root crops).

SEED BED PREPARATION

Cultivation for seed bed preparation has two contrasting effects on weeds: (i) it eliminates the emerged vegetation resulting from after primary tillage; and (ii) it stimulates weed seed germination and consequent seedling emergence, thanks to soil mixing and reallocation of seeds towards shallower soil layers. Together, these two effects can be exploited through application of the false (stale) seed bed technique, a preventive method with the specific aim of reducing weed emergence in the next crop cycle.

The false seed bed technique consists in the anticipation of cultivation time for seed bed preparation, in order to stimulate as much as possible the emergence of weeds prior to sowing. Emerged weeds are then destroyed by the next cultivator pass or by application of a total herbicide (e.g. glyphosate), the latter being useful especially where perennial weeds are present. At sowing time, the seed bank of those weed species able to emerge together with the crop is then already partially depleted and their emergence in the crop is reduced. Cultivation can be performed with any mechanical tools, but spring-tine harrows (Figure 3) are preferable because of their high working capacity and relatively low cost. Application of the false seed bed technique can reduce weed emergence > 80 percent compared to standard seed bed preparation (van der Weide et al. 2002). Obviously, application of this technique implies that there should be enough time (at least 2-3 months in temperate climates) between harvest of the previous crop and sowing of the next crop to allow weeds to emerge. For this method to be effective, the soil must have enough moisture to sustain weed-seed germination. Therefore, this method is useless where soil water availability is limited. Where farmers expect high rainfall events to occur between primary tillage and crop sowing, they must evaluate whether anticipation of seed bed preparation could increase the risk of damaging soil structure or delaying crop sowing because the soil cannot be timely worked: both effects may counteract the benefits of the false seed bed technique and therefore require careful evaluation.

Table 2. Relative density (percentage) of the main weed species and total weed density (plants m^-2) observed in winter wheat just prior to post-emergence herbicide application (data pooled over two years and three nitrogen rates). Data are shown as arcsine-transformed (relative density) or square root-transformed (total weed density) means to allow direct interpretation of SEDs. DP = deep ploughing (50 cm); SP = shallow ploughing (25 cm); TLP = two-layer ploughing (shallow ploughing + subsoiling at 50 cm); CP = chisel ploughing (50 cm); MT = minimum tillage (rotary harrowing at 15 cm). Significance of the F tests (indicated as superscript): *,**,***, ns = P £ 0.05, 0.01, 0.001, and non significant, respectively. After Bàrberi et al. (2001), modified.

Figure 3. A spring-tine harrow with four modular frames of 1.5 m width each that can be adjusted independently

(Photo: P. Bàrberi).

SOIL SOLARIZATION

Soil solarization is a preventive method that exploits solar heating to kill weed seeds and therefore reduce weed emergence. This method is only briefly explained here because it is exhaustively treated in another chapter of this publication.

High soil temperature, if lasting long enough, is able to kill the reproduction structures of pests, diseases, and weeds. Solarization can be defined as a soil disinfection method that exploits the solar energy available during the warmest period of the year. To increase the solarization effect as much as possible, the soil surface must be smooth and must contain enough water to favour heat transfer down the profile and to make reproductive structure of pests, diseases and weeds more sensitive to heat damage. For this reason, prior to solarization the soil is usually irrigated and a plastic mulch film is laid down onto the soil to further increase soil heating and to avoid heat dissipation to the atmosphere.

The success of soil solarization as a weed control method does not depend on the actual value of peak temperature reached in the soil but rather on temperature duration above a certain threshold (45°C) on a daily basis (Horowitz et al. 1983). It follows that soil solarization can only be used in warm climates or under glasshouse conditions in warm-temperate and Mediterranean climates. For example, a significant reduction in weed emergence was observed over the following 12 months after one-month’s solarization in a tunnel glasshouse used for vegetable production in Central Italy (Temperini et al. 1998). To retain as much as possible the weed control effect of solarization, the soil must not be cultivated subsequently because otherwise weed seeds present in deeper soil layers (less affected by heating) are brought up to the soil surface and can germinate.

MANAGEMENT OF DRAINAGE AND IRRIGATION SYSTEMS

Careful choice and maintenance of drainage and irrigation systems is an important preventive measure to reduce on-field weed infestation. Periodical clearance of weed vegetation established along ditches prevents it from invading the field. Where it is economically feasible, substitution of ditches with subterranean drains eliminates a potential source of weed infestation. Use of localized (e.g. trickle) irrigation systems favour crop development to the detriment of weeds (Berkowitz, 1988). In contrast, broadcast irrigation systems often favour weeds because most of them have a higher water use efficiency (dry biomass production per unit water used for evapotranspiration) than the crop.

CROP RESIDUE MANAGEMENT

Cultivation of crop residues stimulates weed seed germination and emergence and hence is positive because it depletes their seed bank. However, care should be taken to prevent the emerged weeds from setting seeds, thus replenishing soil seed reserves. Stubble cultivation can be negative in environments characterized by high mineralization rates of soil organic matter. In these cases, it is better not to disturb the soil or to chop residues and distribute them as evenly as possible over the soil surface to smother weeds germinating in the understorey. This is actually the same effect that can be expected from cover crops when they are used as a dead mulch.

Although seeds of many weed species can be devitalized by stubble burning, this technique is always to discourage because of its negative effect on soil organic matter content.

CULTURAL METHODS

Crop sowing time and spatial arrangement

In some cases, modification of crop sowing date, density and pattern can reduce weed emergence and/or increase crop competitive ability (Mohler, 1996), although this effect is very much dependent on crop species and environment. Spandl et al. (1998) observed that, compared to autumn-sown wheat, control of Setaria viridis in the spring-sown cereal was favoured because the weed emerged in a single flush instead of several flushes, thus being more vulnerable to direct weed control methods (herbicides or cultivation). In cases like this, the crop sowing date can be used by the farmer as a cultural weed management method. In other crops (e.g. vining pea and potato), an increase in seeding rate may turn into higher competitive ability against weeds, but this is often to the detriment of yield because of higher intra-specific competition between pea plants (Lawson and Topham, 1985), or decreased tuber quality and increased potato susceptibility to diseases (Litterick et al. 1999).

In contrast, for crops showing higher phenotypic plasticity, modification of seeding rates and/or pattern may have better chances of being exploited in weed management strategies. This may be the case of pigeon bean (Vicia faba var. minor), a legume suited to Mediterranean environments that is both a valuable protein source for animal feeding and a soil fertility-building crop. Pigeon bean can be sown either in narrowly-spaced rows (ca. 15 cm) or in widely-spaced rows (40-70 cm). In the first case, pod number and grain yield per plant decrease and height of pod insertion on the stem increases (which reduces yield losses resulting from mechanical harvest), but grain yield per unit area and seed crude protein content are still good (Bonari and Macchia, 1975). Thanks to this phenotypic plasticity, it is likely that the spatial arrangement of this crop may be optimized further, for example, by sowing it in paired rows and using an interrow distance (ca 40 to 50 cm) that allows hoeing between the rows, thus probably achieving higher weed control.

Use of transplants instead of seeds (e.g. in vegetable crops) also increases crop competitive ability because it increases the differential of development between crop and weeds to the advantage of the former. Additionally, use of transplants can increase the selectivity (i.e. the ratio between damage to weeds and to crop) of torsion weeders (Figure 4), that are simple and cheap mechanical tools for intra-row weed control (Melander, 2000). In sugar beet, for example, mechanical weed control can already be performed five days after transplanting, with little crop damage. A possible negative side-effect of the use of sugar beet transplants is the higher incidence of root forking, which can decrease produce quality. Compared to direct sowing, the use of transplants often increases crop production costs and labour requirements.

Figure 4. Torsion weeders, i.e. spring tines that can be coupled with hoeing tools to allow intra-row weed control

(Photo: P. Bàrberi).

Crop genotype choice

Different genotypes of the same crop possess traits that may turn into a higher or lower competitive ability against weeds. These traits are typically those related to faster seedling emergence, quick canopy establishment (Rasmussen and Rasmussen, 2000), and higher growth rates in the early stages. Use of these genotypes can therefore reduce the need for direct weed control measures (e.g. herbicides or cultivation).

The potential for selecting crop genotypes with competitive traits has been demonstrated in Australian wheat accessions, although the expression of competitive advantage in a field situation is strongly influenced by environmental conditions (Lemerle et al. 2001).

Not all traits that give crops a competitive advantage against weeds can be exploited; for example, plant height, which is usually correlated with weed suppression (Benvenuti and Macchia, 2000), is often negatively correlated with crop yield and positively correlated with sensitivity to lodging.

Higher genotype competitive ability can also be related to the production and release of allelochemicals that inhibit weed emergence and growth. Olofsdotter (2001) showed that some rice varieties are able to exert a considerable allelopathic activity against weeds, therefore there is potential for using crop genotype choice as a cultural method for weed management in rice.

Cover crops (used as living mulches)

Cover crops can also be used as a living mulch, i.e. they can be grown together with a cash crop, usually in alternate rows. In this case, cover crop benefits are mainly related to enhanced weed suppression and soil moisture conservation. However, it is very important to avoid competition between the cash crop and the living mulch. In this respect, growth of the living mulch needs to be taken constantly under control with mowing or sub-lethal doses of herbicides, to avoid the living mulch overcompeting with the cash crop (Figure 5). In this respect, living mulch management is not easy, and the convenience of this method remains doubtful in environments where competition for light or water can be substantial.

Figure 5. Living mulch between sugar beet and subterranean clover (Trifolium subterraneaum). Top: good living mulch development, bottom: excessive living mulch development

(Photo: P. Bàrberi).

Intercropping

Another cultural method for increasing crop competitive ability against weeds is intercropping. Like cover crops, intercrops increase the ecological diversity in a field. Additionally, they increase the use of natural resources by the canopy and, compared to sole crops, they often compete better with weeds for light, water and nutrients (Liebman and Dyck, 1993). For example, compared to sole cropping, a leek-celery intercrop sown in a row-by-row layout decreased relative soil cover of weeds by 41 percent, reduced the density and biomass of Senecio vulgaris by 58 percent and 98 percent respectively, and increased total crop yield by 10 percent (Baumann et al. 2000). Increased weed suppression and crop yield has also been demonstrated in many environments for cereal-legume intercrops (Ofori and Stern, 1987). As in the case of living mulching, the success of intercropping relies on the best match between the requirements of component species for light, water and nutrients, which increases resource use complementarity and reduces competition between the intercrops. In practice, this means optimizing intercrop spatial arrangement, relative plant densities and crop relative growth over time in any given environment (Ofori and Stern, 1987).

Fertilization

Pre-sowing N fertilization can increase crop competitive ability against weeds in crops having high growth rates at early stages, but this effect is modulated by the type of weeds prevailing in a field. For example, in sunflower grown in Mediterranean conditions, a pre-sowing application of synthetic N fertilizer increased the suppression of late-emerging weeds like Chenopodium album, Solanum nigrum and Xanthium strumarium compared to split application, i.e. 50 percent pre-sowing and 50 percent top-dressing (Paolini et al. 1998). In contrast, the same technique resulted in a competitive advantage for early-emerging weeds like Sinapis arvensis. Similarly, anticipation or delay of top-dressing N application in sugar beet increased crop competitive ability with dominance of late- or early-emerging weeds respectively (Paolini et al. 1999).

Modulation of crop-weed competitive interactions through crop fertilization is unlikely to be feasible when organic fertilizers or amendments (e.g. manure) are used, because of the slow release of nutrients from these sources. However, application of fertilizers (either synthetic or organic) along with, or in close proximity to the crop row, can improve weed management because it increases the relative chances of the crop to capture nutrients (especially N) to the detriment of weeds (Rasmussen, 2000).

CONCLUSIONS

Farmers have several preventive and cultural methods in their arsenal that they can put together to build up a good weed management strategy. The convenience of using one method instead of another depends on local attitudes and constraints such as availability of money and labour, access to technical means (e.g. seeds, fertilizers, herbicides), environmental, social and economic features that may limit the range of feasible agronomic choices (e.g. length of the growing season, rainfall and temperature patterns, soil mineralization rate, farm and market structure, cultural heritage, existence of advisory services, etc.). However, highest diversification of the cropping system (i.e. crop sequence and associated cultural practices) based on agro-ecological principles is the key to effective long-term weed management in any situations. In this respect, the systematic inclusion of preventive and cultural methods for weed management must always be pursued. This obviously implies that farmers must be educated to acquire a higher level of knowledge and technical skills. Simple solutions, such as monocropping and reliance on herbicides as the only direct weed-control method may be successful in the short term, but are never rewarding in the long term.

BIBLIOGRAPHY

Bàrberi, P., Bonari, E. & Mazzoncini, M. 2001. Weed density and composition in winter wheat as influenced by tillage systems. In Proc. I^st World Congress on Conservation Agriculture, Madrid (Spain), 1-5 October, 451-455.

Bàrberi, P. & Lo Cascio, B. 2001. Long-term tillage and crop rotation effects on weed seedbank size and composition. Weed Res. 41: 325-340.

Bàrberi, P. & Mazzoncini, M. 2001. Changes in weed community composition as influenced by cover crop and management system in continuous corn. Weed Sci. 49: 491-499.

Baumann, D.T., Kropff, M.J. & Bastiaans, L. 2000. Intercropping leeks to suppress weeds. Weed Res. 40: 359-374.

Benvenuti, S. & Macchia, M. 2000. Role of durum wheat (Triticum durum Desf.) canopy height on Sinapis arvensis L. growth and seed production. In Proc. XIème Colloque Intl. sur la Biologie des Mauvaises Herbes, Dijon, France. pp.305-312.

Berkowitz, A.R. 1988. Competition for resources in weed-crop mixtures. In Weed Management in Agroecosystems: Ecological Approaches. eds. Altieri, M.A. & Liebman, M. 89-119, CRC Press, Boca Raton, Florida, USA.

Blum, U., King, L.D., Gerig, T.M., Lehman, M.E. & Worsham, A.D. 1997. Effects of clover and small grain cover crops and tillage techniques on seedling emergence of some dicotyledonous weed species. American J. of Alternative Agriculture 4: 146-161.

Bonari E. & Macchia, M. 1975. Effetto dell’investimento sulla produzione del favino (Vicia faba L. var. minor Peterm Beck). Rivista di Agronomia 9: 416-423 (in Italian with English abstract).

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

Cousens, R. & Mortimer, M. 1995. Dynamics of weed populations. Cambridge University Press, Cambridge, UK. 332 pp.

Froud-Williams, R.J. 1983. The influence of straw disposal and cultivation regime on the population dynamics of Bromus sterilis. Annals of Applied Biology 103: 139-148.

Froud-Williams, R.J. 1988. Changes in weed flora with different tillage and agronomic management systems. In Weed Management in Agro-ecosystems: Ecological Approaches, eds. Altieri, M.A. & Liebman, M., 213-236, CRC Press, Boca Raton, Florida, USA.

Gallandt E.R., Libeman, M. & Huggins, D.R. 1999. Improving soil quality: implications for weed management. In Expanding the Context of Weed Management, ed. Buhler, D.D. 95-121. The Haworth Press, New York, USA.

Horowitz, M., Regev, Y. & Herzlinger, G. 1983. Solarization for weed control. Weed Sci. 31: 170-179.

Jordan, N.R., Zhang, J. & Huerd, S. 2000. Arbuscular-mychorrhizal fungi: potential roles in weed management. Weed Res. 40: 397-410.

Karlen, D.L., Varvel, G.E., Bullock, D.G. & Cruse, R.M. 1994. Crop rotations for the 21th century. Advances in Agronomy 53: 1-45.

Lal, R., Regnier, E., Eckert, D.J., Edwards, W.M. & Hammond, R. 1991. Expectations of cover crops for sustainable agriculture. In Cover Crops for Clean Water, ed. Hargrove, W.L., pp.1-11. Soil and Water Conservation Society Publication, Ankey, USA.

Lawson, H.M. & Topham, P.B. 1985. Competition between annual weeds and vining peas grown at a range of population densities: effects on the weeds. Weed Res. 25: 221-229.

Lègère, A., Samson, N. & Rioux, R. 1993. Perennial weeds in conservation tillage systems: more of an issue than in conventional tillage systems? In Proc. 1993 Brighton Conference - Weeds, 747-752.

Lemerle, D., Verbeek, B. & Orchard, B. 2001. Ranking the ability of wheat varieties to compete with Lolium rigidum. Weed Res. 41: 197-209.

Liebman, M. & Davis, A.S. 2000. Integration of soil, crop, and weed management in low-external-input farming systems. Weed Res. 40: 27-47.

Liebman, M., Drummond, F.A., Corson, S. & Zhang, J. 1996. Tillage and rotation crop effects on weed dynamics in potato production systems. Agronomy J. 88: 18-26.

Liebman, M. & Dyck, E. 1993. Crop rotation and intercropping strategies for weed management. Ecological Applications 3: 92-122.

Litterick, A.M., Redpath, J., Seel, W. & Leifert, C. 1999. An evaluation of weed control strategies for large-scale organic potato production in the UK. In Proc. 1999 Brighton Conference - Weeds, Brighton, UK. 951-956.

Melander, B. 2000. Mechanical weed control in transplanted sugar beet. In Proc. of the 4^th Workshop of the EWRS Working Group on Physical and Cultural Weed Control, Elspeet The Netherlands, 25.

Mohler, C.L. 1996. Ecological bases for the cultural control of annual weeds. J.of Production Agriculture 9: 468-474.

Mohler, C.L. & Teasdale, J.R. 1993 Response of weed emergence to rate of Vicia villosa Roth and Secale cereale L. residue. Weed Res. 33: 487-499.

Ofori, F. & Stern, W.R. 1987. Cereal-legume intercropping systems. Advances in Agronomy 41: 41-90.

Olofsdotter M. 2001. Rice - a step toward use of allelopathy. Agronomy J. 93: 3-8.

Paolini, R., Principi, M., Del Puglia, S. & Lazzeri, L. 1998. Competitive effects between sunflower and six broad-leaved weeds. In Proc. 6^th EWRS Mediterranean Symposium, Montpellier, France, 81-88.

Paolini, R., Principi, M., Froud-Williams, R.J., Del Puglia S. & Biancardi, E. 1999. Competition between sugarbeet and Sinapis arvensis and Chenopodium album, as affected by timing of nitrogen fertilization. Weed Res. 39: 425-440.

Rasmussen K. 2000. Can slurry injection improve the selectivity of weed harrowing in cereals? In Proc. of the 4^th Workshop of the EWRS Working Group on Physical and Cultural Weed Control, Elspeet, The Netherlands, 33-34.

Rasmussen, K. & Rasmussen, J. 2000. Barley seed vigour and mechanical weed control. Weed Res. 40: 219-230.

Spandl E., Durgan, B.R. & Forcella, F. 1998. Tillage and planting date influence foxtail (Setaria spp.) emergence in continuous spring wheat (Triticum aestivum). Weed Tech.12, 223-229.

Spandl, E., Durgan, B.R. & Forcella F. 1999. Foxtail (Setaria spp.) seedling dynamics in spring wheat (Triticum aestivum) are influenced by seeding date and tillage regime. Weed Sci. 47: 156-160.

Teasdale, J.R. & Mohler, C.L. 1993. Light transmittance, soil temperature, and soil moisture under residue of hairy vetch and rye. Agronomy J. 85: 673-680.

Teasdale, J.R & Mohler, C.L. 2000. The quantitative relationship between weed emergence and the physical properties of mulches. Weed Sci. 48: 385-392.

Temperini, O., Bàrberi, P., Paolini, R., Campiglia, E., Marucci, A. & Saccardo, F. 1998. Solarizzazione del terreno in serra-tunnel: effetto sulle infestanti in coltivazione sequenziale di lattuga, ravanello, rucola e pomodoro. In Proc. XI SIRFI Biennial Congress, Bari, Italy. 12-13 November, 213-228 (Italian, with English abstract).

Van Der Weide, R.Y., Bleeker, P.O. & Lotz, L.A.P. 2002 Simple innovations to improve the effect of the false seed bed technique. In Proc. of the 5^th Workshop of the EWRS Working Group on Physical and Cultural Weed Control, Pisa, Italy. 3-4.

Zanin, G., Otto, S., Riello L. & Borin, M. 1997 Ecological interpretation of weed flora dynamics under different tillage systems. Agriculture Ecosystems & Environment 66: 177-188.