No description available.

Les plantes envahissantes menacent la biodiversité ainsi que les activités humaines. Afin de les maîtriser, la pulvérisation d'herbicides est une méthode fréquemment employée en Amérique du Nord. Cette approche ne fait pas toujours consensus et est même parfois interdite ou restreinte, ce qui justifie le recours à d'autres options. Les alternatives peuvent toutefois s'avérer rares, comporter d'importantes limitations ou sont peu documentées. Cette étude vise à tester l’efficacité de méthodes permettant de maîtriser et de prévenir les invasions de roseau commun (Phragmites australis), l'une des plantes envahissantes les plus problématiques sur le continent nord-américain, tout en limitant au minimum l'utilisation d'herbicides. Le potentiel de quatre méthodes de lutte aux petites populations de roseau bien établies a d'abord été évalué : l’excavation avec enfouissement sur place, le bâchage, l’excavation avec enfouissement sur place combinée au bâchage, et la fauche répétée. Les résultats ont montré que l'excavation avec enfouissement sur place, avec ou sans bâchage, a entraîné une élimination presque totale des populations visées, ce qui est comparable ou supérieur à l'effet généralement obtenu avec la pulvérisation d'herbicide. Le bâchage avec des toiles opaques, maintenues pendant un an, a pour sa part entraîné une maîtrise partielle du roseau, suggérant qu'une application prolongée serait nécessaire pour l'éradication de la plante. La fauche répétée, exécutée à raison de cinq fauches par été pendant deux ans, a fourni une efficacité mitigée. Les résultats suggèrent néanmoins que la fauche pendant plusieurs années contribue à affaiblir la plante, ce qui pourrait aider à son confinement. Une méthode additionnelle a été expérimentée afin de traiter les tiges éparses de roseau tout en limitant les risques d'effets hors cibles, soit le badigeonnage manuel d’herbicide. Suite à ces tests, les résultats ont montré une diminution importante de la densité des tiges, ce qui suggère que la méthode est efficace afin d'éliminer les repousses après un traitement initial, et pourrait également être employée sur de jeunes populations clairsemées. L'effet d'un ensemencement préventif de plantes herbacées sur l'établissement de semis de roseau a également été étudié, suite à des traitements sur de vastes parcelles de sol nu. Les résultats suggèrent que la méthode est efficace afin de limiter la propagation du roseau par semences et qu'un suivi périodique suite à l'intervention serait suffisant afin de maintenir l'effet préventif. / Invasive plants threaten biodiversity as well as human activities and in North America, attempts to control them often rely on the spraying of herbicides. This approach does not always achieve consensus and is sometimes forbidden or restricted, encouraging managers to consider other options. Alternatives can however prove to be rare, involve serious limitations or be poorly documented. This study aims at testing the efficiency of methods for the control and prevention of invasions by common reed (Phragmites australis), one of the most problematic invasive plants on the North American continent, all while reducing herbicide use to a minimum. The potential of four methods for the control of small, well established stands of common reed has first been evaluated: excavation and on-site burial, covering with heavy-duty tarps, excavation and on-site burial combined with covering with heavy-duty tarps, and repeated cutting. Results have shown that excavation with on-site burial, followed or not by covering with heavy-duty tarps, led to an almost complete elimination of the populations targeted, which is comparable or superior to the effect typically obtained with herbicide spraying. Covering common reed with opaque heavy-duty tarps, held in place for one year, lead to partial control of the plant, suggesting that holding the tarps in place for a longer period would be necessary to foresee the eradication. Cutting, repeated five times per summer for two years, gave mixed results. The results suggest, however, that cutting during multiple years contributes to weakening the plant, which could help to its confinement. An additional method, the hand-wicking of common reed stems with herbicide, has been tested to treat scattered stems while limiting off-target effects. Following these tests, the results have shown an important reduction of the density of reed stems, which suggests that the method is effective to control common reed regrowth after an initial treatment, and could be used on young stands. The effect of seeding herbaceous plants on the establishment of common reed seedlings has also been evaluated, following the sowing of vast plots of bare soil. The results suggest that the method is an efficient way of limiting the propagation of common reed through seed dispersal. Results also suggest that periodic follow-ups to eliminate seedlings would be sufficient to maintain the preventive effect.

Made available in DSpace on 2017-07-25T19:29:38Z (GMT). No. of bitstreams: 1 JoseAlfredoBSantos.pdf: 575338 bytes, checksum: 6d87ba82656e81978e4b945304868267 (MD5) Previous issue date: 2008-06-26 / With the objective of evaluating alternatives for oval-leaf false button weed management in black beans crops for familiar farmers of the Southern region of Paraná State, studies were carried out, in Ponta Grossa at Iapar Experimental Station and in Irati at a farmer field. A survey was also carried out to evaluated the management practices of oval-leaf false buttonweed by the farmers. One study in Ponta Grossa had as objective to evaluate the period of emergence of oval-leaf false buttonweed during one year, in area with and without vegetal covering, and also with and without mobilization of the soil. In another study also in Ponta Grossa an experiment was installed to evaluate the influence of cover crops in the reduction of the infestation of buttonweed in no-till with three winter cover crops: black oat, radish, black oat with common vetch and a fallow treatment. The experiment in the region of Irati had as objective to evaluate, besides the influences of cover crops, also the liming in the reduction of the population of buttonweed in black beans crop. The cover crops were the same as the experiment in Ponta Grossa. The management practices was evaluated through a questionnaire applied to seventeen farmers separated in two groups: 1) those who still regarded false buttonweed as a problem in the farm, and 2) those farmers who had false buttonweed as a problem but were able to solve it. In the study of period of emergence of false buttonweed the main peaks of flush were in October, November and December. The experiments to evaluate the effect of winter cover crops in Ponta Grossa and Irati have shown no effect on false buttonweed populations overall. There was also not diference in the liming treatments in Irati. The survey with the farmers of the region has shown that successfull farmers in the management of false buttonweed were those who use integrated weed management strategies. / Com objetivo de avaliar alternativas de manejo da erva-quente (Spermacoce latifolia Aubl) em cultivo de feijão, foram realizados estudos na região Centro-Sul do Paraná, por meio da instalação de experimentos em Ponta Grossa e em Irati; também foi feito um diagnóstico para verificar os procedimentos adotados pelos agricultores da região no manejo dessa espécie. Em Ponta Grossa, um estudo teve como objetivo avaliar o fluxo de emergência de erva-quente durante um ano, em área com e sem cobertura vegetal, e com e sem mobilização do solo. Num segundo estudo, em Ponta Grossa, avaliou-se a influência de plantas de cobertura na redução da infestação de erva-quente em plantio direto, utilizando-se três espécies de cobertura vegetal de inverno: aveia-preta, nabo-forrageiro, aveia-preta com ervilhaca-comum e pousio. O experimento de Irati teve como objetivo avaliar a influência de plantas de cobertura e também a correção do solo na redução da população de erva-quente em cultivo sucessivo de feijão. As coberturas de inverno foram as mesmas usadas no experimento de Ponta Grossa. No levantamento das práticas de controle na região foi elaborado um questionário, que foi utilizado junto a dezessete agricultores nos municípios de Mallet, Rio Azul, Irati e Prudentópolis, separados em dois grupos: 1) os que ainda consideravam a erva-quente um problema na propriedade e 2) aqueles que já tiveram a erva-quente como problema e conseguiram solucioná-lo. No estudo do fluxo de emergência verificou-se que os principais meses de pico de emergência de erva-quente foram em outubro, novembro e dezembro. No conjunto dos experimentos para estudo da influência de plantas de cobertura em Ponta Grossa e Irati não se verificou efeito das coberturas testadas na população da erva-quente. Também não houve diferença quando se avaliou a influência do calcário na infestação de erva-quente. No levantamento das práticas de manejo da erva-quente se verificou que o grupo de agricultores com sucesso no controle da erva-quente é o que utiliza um sistema de rotação de culturas e manejo integrado de plantas daninhas.

La lutte biologique par conservation des habitats est une voie pour réduire la dépendance et l’impact des traitements insecticides qui sont nombreux et toxiques. Elle est particulièrement appropriée pour les vergers d’agrumes qui comprennent, souvent, de grandes surfaces non cultivées. Les auxiliaires présents en plus grand nombre dans un enherbement sont alors plus facilement mobilisables lors d’attaques massives de ravageurs sur les arbres. Ce travail de thèse a pour objectif d’évaluer l’impact des modes de gestion de l’enherbement (fauche, broyage, travail du sol et herbicide) sur les communautés végétales et les populations d’auxiliaires au sein du verger, en s’intéressant à deux communautés évoluant à deux échelles spatiales distinctes : les Phytoseiidae et les Coccinellidae. Nos résultats ont montré une différenciation fonctionnelle des communautés végétales soumises à différents modes de gestion dans le temps. Les Phytoseiidae sont présents en abondances comparables dans toutes ces communautés végétales mais sont cependant influencés par l’abondance dans le couvert des espèces végétales portant des densités de trichomes modérés et des domaties. De même, les abondances de Coccinellidae sont positivement corrélées à l’abondance des Poaceae sur la parcelle alors que les modalités de gestion de l’enherbement testées n’ont pas montré d’effet significatif. Ces connaissances ont pu être mobilisées et articulées pour l’élaboration d’un prototype d’outil d’aide à la décision contribuant à la conception de vergers agroécologiques. Il permet l’évaluation ex ante de scénarios de pratiques de gestion de l’enherbement en vue d’optimiser la lutte biologique par conservation au sein des vergers d’agrumes. Nous avons choisi une approche fonctionnelle qui permettra de généraliser facilement l’outil à d’autres cultures moyennant un élargissement de la base de données comportant les traits fonctionnels aux espèces végétales non renseignées et une bonne connaissance des affinités des auxiliaires que l’on cherche à favoriser / Biological control through habitat conservation is a way to reduce the reliance on and the impact of insecticide treatments that are numerous and toxic. It is particularly suitable for citrus orchards which include large areas often uncultivated. The natural enemies present in greater number in a vegetal cover are then more easily mobilized during massive attacks of pests on trees. The aim of this thesis is to evaluate the impact of different weed management methods (mowing, milling, tillage and herbicide) on plant communities and natural enemies populations within the orchard, focusing on two communities that evolve at two distinct spatial scales: the Phytoseiidae and the Coccinellidae. Our results showed a functional differentiation of vegetal communities managed with different weed management methods over time. Phytoseiidae are present in comparable abundances in all these plant communities but are influenced by the abundance of plant species with densities of moderate trichomes and domaties in the vegetal cover. Similarly, the abundances of Coccinellidae are positively correlated with the abundance of Poaceae species on the plot whereas the weed management modalities have no significant impact. This knowledge has been mobilized and articulated for the development of a decision-making tool prototype contributing to the design of agro-ecological orchards. It allows ex ante evaluation of scenarios of weed management practices in order to optimize biological control by conservation in citrus orchards. We have chosen a functional approach which will easily allow to generalize the tool to other crops by expanding the database with functional traits to non-informed plant species and a good knowledge of the affinities of natural enemies to promote

På Sveriges järnvägsnät bekämpas ogräs med det kemiska bekämpningsmedlet glyfosat. En omställning är nu på gång och fler bekämpningsmetoder behöver introduceras, bland annat för att motverka resistensutveckling hos ogräsen. I detta arbete jämfördes bekämpning med pelargonsyra, hetvatten, vattenånga och elektricitet med avseende på effektivitet, kostnad, energi- och vattenåtgång. Glyfosat användes som referens. Jämförelsen gjordes genom att sammanställa tillgängliga forskningsrapporter. Bekämpningsmetodernas användbarhet analyserades genom att uppskatta resursåtgången för att ersätta den kemiska bekämpningen på linjen eller driftplatser, eller för att bekämpa restriktionsytor. Det undersöktes också huruvida kostnaden för ogräsbekämpning över tid har betydelse för val av ballastfri järnväg. Resultatet visade att de alternativa metoderna är både dyrare och kräver en högre energi- och vattenåtgång än glyfosat. De ger också en mindre effektiv ogräsbekämpning och upprepade behandlingar behöver därför göras under växtsäsongen. Detta i kombination med att de med dagens teknik appliceras i låg hastighet gör bekämpningen dyr och tidskrävande. För att kunna säga i vilken omfattning de alternativa bekämpningsmetoderna är möjliga att implementera behövs riktlinjer för vilken kostnad, effektivitet samt vatten- och energiåtgång som kan tolereras. Resultatet tyder på att det kan vara möjligt att införa alternativa bekämpningsmetoder på driftplatser och restriktionsytor. För att ersätta glyfosat längs hela järnvägslinjen behöver de utvecklas, speciellt med avseende på effektivitet. För bekämpning med hetvatten och vattenånga var den stora vattenåtgången en stor nackdel. Bekämpningsmetodernas kostnad utgjorde under 5 % av det totala underhållsbehovet på en ny järnväg av ballast. Den ökade kostnaden för ogräsbekämpning ledde som mest till att en ballastfri järnväg blev lönsam två år tidigare för de alternativa metoderna än för glyfosat. Valet mellan en ballasterad och ballastfri järnväg bedömdes inte påverkas nämnvärt av detta. Dock identifierades faktorer som kan göra bekämpningskostnaden högre än vad den här rapporten har kunnat påvisa. Vid en omställning från glyfosat till alternativa metoder bör man därför vid lönsamhetsberäkningar för ny järnväg vara vaksam på vad bekämpningen på det ballasterade alternativet skulle kosta. / Weeds are growing along the Swedish railway network. In order for them to not cause problems they are being removed chemically with the herbicide glyphosate. Due to problems such as a risk for an increase in weeds that are tolerant to glyphosate alternative methods for weed control are now being considered.  This report compared pelargonic acid, hot water, hot water steam and electricity as control methods in terms of efficacy, cost, energy and water consumption. Glyphosate was used as a reference. The comparison was made by reviewing available research articles. The usability of the control methods was then analysed by estimating the use of resources necessary to implement each method for three different scenarios. Whether the choice between a ballasted or fixed rail was affected or not by the cost for weed control was also investigated. The results showed that the alternative weed control methods required more energy and water than glyphosate. They also provide less effective weed control which means repeated treatments are necessary during the growing season. This, combined with a low application speed, makes the weed control expensive and time consuming. To be able to say to what extent each control method is possible to implement guidelines are needed for tolerated cost, effectiveness, energy- and water consumption. The result suggests that it is possible to introduce alternative control methods to a certain extent. However, in order to replace glyphosate along the entire main line, the methods all need to be developed, especially in terms of effectiveness. For hot water and steam treatment there is also a need for a decrease in water consumption.  The cost for weed control accounted for less than 5 % of the total cost for maintenance on a new ballasted railway. The rise in cost due to alternative weed management resulted in the fixed track becoming profitable up to two years earlier with the alternative management compared to glyphosate. The choice between a ballasted and fixed track was not considered to be significantly affected by this. However, factors were identified that could make the cost of weed control higher than this report showed. Therefore, when alternative methods are introduced, one should consider the cost for weed control on the ballasted track when making the choice between the two types of tracks.

Crop rotation may be used to prevent the continuous selection of particular weed species adapted to one crop type. This might be useful for weed management, economy in herbicide applications and promoting biodiversity. Common simple crop sequences might be diversified by introducing perennial forage crops. Impacts of such perennial crops on weeds were studied with four approaches : 1) Large-scale weed surveys in 632 fields in western France showed that weed species composition differed most strongly between perennial alfalfa crops and annual crops. Comparisons of fields before, during and after perennial alfalfa suggested that community composition varies in a cyclic way during such crop rotations. Several weed species problematic in annual crops were suppressed during and after perennial crops, but the appearance of other species led to equal or even higher plant diversities. 2) A 3-year field experiment with contrasting crop management options allowed an investigation of the underlying mechanisms for this: The absence of soil tillage reduced weed emergence but increased the survival of established plants. The permanent vegetation cover and frequent hay cuttings reduced weed growth, plant survival and seed production. 3) Greenhouse experiments testing the regrowth ability of individual plants after cutting showed strong differences between species and functional groups. An two-factorial experiment suggested that the negative impacts of cutting and competition on weed growth were mainly additive. 4) Special measurements of weed seed predation in the field experiment showed positive correlations with vegetation cover, indicating that this ecosystem service may be particularly fostered by perennial crops. Consistent preferences of seed predators for certain weed species indicates that seed predation may be another cause of the observed weed community shifts.

Weed management in sustainable farming systems often requires the integration of several different pest management techniques. Cultural, biological, and chemical pest control efforts can be utilized to contribute to the common goal of reducing weeds in vegetable row crop systems. This research addresses how common disturbances such as tillage and insecticide use affect: carabid beetle assemblages; predation of weed seeds by invertebrates; and weed seed recruitment. Field experiments were conducted over three years at two different locations on the OSU Vegetable Research Farm, Linn Co., OR. We found that activity-density of carabid beetles varied seasonally and peaked in late August-September each year. Insecticide applied in year 1 affected seed loss in year 2, suggesting possible long-term effects of land management on weed seed removal. Weed recruitment was highly variable between treatment, site, and year. Conserving biological weed control agents in combination with cultural techniques such as reducing tillage and the use of cover crops, helps growers shift from expensive, density-independent control efforts to more ecological, long-term solutions for weed management in agroecosystems. / Graduation date: 2011 / Access restricted to the OSU community, at author's request, from Dec. 15, 2010 - Dec. 15, 2011.

The spider mite occurrence, on soybean (Glycine max (L.) Merril), is sporadic and associated whit dry periods. Recently, in continuous years, was reported mite outbreaks in soybean, in some sites of Rio Grande do Sul State. In this context two research was conducted, one in summer crop 2004/05 and other in 2005/06. The first was carried for study the mite occurrence and geographic distribution, on soybean, from some place of Rio Grande do Sul State. The second was conducted for study the effect of the two soybean cultivars and of some weed management systems on mite density and study their population dynamic on soybean. In summer crop 2004/05 was carried some mite samplings in soybean field at State. The mite species found, all beloung to the Tetranychidae family, are Mononychellus planki (McGregor), Tetranychus desertorum Banks, T. gigas Pritchard & Baker, T. ludeni Zacher and T. urticae Koch. Most of this species are well distributed in the diferents sampling regions of this study. From the results was made a species distribution map and a taxonomic key. In summer crop 2005/06, in Santa Maria, RS, 11 treatments was study. Its was constitute in two soybean cultivars, one transgenic glyphosato-tolerant and other non-transgenic, every one whit five weed management systems: (1) without management; (2) hand management; (3) herbicide bentazone + sethoxydim; (4) bentazone only; (5) sethoxydim only. The 11th treatment was the herbicide glyphosate sprayed on the transgenic soybean. The mite species found in this study are M. planki and T. gigas. The weed management systems is more determinant, on the spider mite population on soybean, than the soybean cultivar. The mite population presenteded inverse correlation whit weed infestation and weed density. The climatic fators determined the mite population flutuation and the population peak occured in the initial period of soybean pods development. / A ocorrência de ácaros-praga em soja (Glycine max (L.) Merril) é esporádica e está, comumente, associada à períodos de estiagem. Recentemente, em safras seguidas, foram observados ataques severos de ácaros à soja em algumas regiões do estado do Rio Grande do Sul. Neste contexto foram realizados dois experimentos, um na safra agrícola 2004/05 e outro em 2005/06. O primeiro teve objetivo de estudar a ocorrência e a distribuição geográfica de ácaros fitófagos, associados à soja, em municípios de seis regiões produtoras do estado do Rio Grande do Sul. O segundo objetivou avaliar o efeito de duas cultivares de soja e de diferentes sistemas de manejo de plantas daninhas sobre a densidade populacional de ácaros em soja e estudar a dinâmica populacional destes ácaros na cultura. Em 2004/05 foram realizadas amostragens de ácaros em lavouras de soja no Estado. As espécies de ácaros encontradas são Mononychellus planki (McGregor), Tetranychus desertorum Banks, T. gigas Pritchard & Baker, T. ludeni Zacher e T. urticae Koch, todos são tetraniquídeos. A maior parte destas espécies estiveram bem distribuídas nas diferentes regiões amostradas neste levantamento. Com base nestes resultados foi elaborado um mapa com a distribuição das espécies e uma chave de identificação. Em 2005/06 foram estudados, em Santa Maria, RS, 11 tratamentos. Estes consistiram em duas cultivares de soja, uma transgênica glifosato-tolerante e a outra não-transgênica, cada uma sob cinco sistemas de manejo de plantas daninhas: (1) sem controle; (2) arranquio manual; (3) manejo tradicional com os herbicidas bentazona e setoxidim; (4) apenas com bentazona; (5) apenas com setoxidim. E adicionalmente um tratamento com o herbicida glifosato sobre a cultivar de soja transgênica. Os ácaros-praga ocorrentes foram M. planki e T. gigas. A densidade populacional destes ácaros em soja foi mais afetada pelos sistemas de manejo de plantas daninhas do que pela cultivar, e esteve inversamente correlacionada com a quantidade e diversidade de plantas daninhas. A flutuação populacional destes ácaros foi influenciada por fatores climáticos e o pico populacional ocorreu na fase de início da formação de legumes da soja.

La rotation de cultures peut être utilisée pour empêcher la sélection continue d’espèces adventices adaptées à un type de culture. Elle pourrait favoriser la gestion des adventices, l’économie d’herbicides et la biodiversité. Les successions de cultures simples d’aujourd’hui pourraient être diversifiées par des cultures fourragères pérennes. Les impacts des ces cultures sur les adventices ont été étudié utilisant quatre approches : 1) Des relevés d’adventices sur 632 champs dans l’ouest de la France ont montré que la composition spécifique varie le plus entre des cultures fourragères pérennes et des cultures annuelles. Une comparaison des champs avant, pendant, et après des cultures fourragères pérennes a suggéré que la composition des communautés varie d’une manière cyclique pendant ces rotations. Plusieurs espèces problématiques dans des cultures annuelles ont été supprimées pendant et après les cultures pérennes, mais l’apparition d’autres espèces a produit une diversité de plantes comparable, voire supérieure. 2) Une expérimentation au champ de trois ans avec des modes de gestion contrastés a permis d’étudier les mécanismes sous-jacents: L’absence de travail du sol a réduit la levée des adventices, mais a augmenté la survie des plantes adultes. Le couvert végétal permanent et les fauches fréquentes ont réduit la croissance, la survie des plantes et la production de graines. 3) Des expérimentations sous serre analysant la croissance poste fauche de plantes individuelles ont montré des différences importantes entre espèces et groupes fonctionnels. Une expérimentation à deux facteurs a suggéré que les impacts négatifs de la fauche et de la compétition sur la croissance des adventices ont été additifs. 4) Des mesures spéciales de prédation de graines d’adventices sur l’expérimentation au champ ont montré des corrélations positives avec le couvert végétal et la prédation, indiquant une importance particulière de ce service écosystémique dans des cultures pérennes. La préférence des graines de certaines espèces montre que la prédation de graines peut être une autre cause des changements de communautés d’adventices. / Crop rotation may be used to prevent the continuous selection of particular weed species adapted to one crop type. This might be useful for weed management, economy in herbicide applications and promoting biodiversity. Common simple crop sequences might be diversified by introducing perennial forage crops. Impacts of such perennial crops on weeds were studied with four approaches : 1) Large-scale weed surveys in 632 fields in western France showed that weed species composition differed most strongly between perennial alfalfa crops and annual crops. Comparisons of fields before, during and after perennial alfalfa suggested that community composition varies in a cyclic way during such crop rotations. Several weed species problematic in annual crops were suppressed during and after perennial crops, but the appearance of other species led to equal or even higher plant diversities. 2) A 3-year field experiment with contrasting crop management options allowed an investigation of the underlying mechanisms for this: The absence of soil tillage reduced weed emergence but increased the survival of established plants. The permanent vegetation cover and frequent hay cuttings reduced weed growth, plant survival and seed production. 3) Greenhouse experiments testing the regrowth ability of individual plants after cutting showed strong differences between species and functional groups. An two-factorial experiment suggested that the negative impacts of cutting and competition on weed growth were mainly additive. 4) Special measurements of weed seed predation in the field experiment showed positive correlations with vegetation cover, indicating that this ecosystem service may be particularly fostered by perennial crops. Consistent preferences of seed predators for certain weed species indicates that seed predation may be another cause of the observed weed community shifts. / Fruchtfolgen können dazu dienen, die kontinuierliche Selektion von Unkrautarten zu verhindern, die an eine bestimmte Kultur angepasst sind. Dies könnte dem Unkrautmanagement, der Einsparung von Herbiziden, und der Biodiversität dienen. Heutige, sehr einfache Furchtfolgen könnten durch mehrjährige Futterkulturen diversifiziert werden. Die Einflüsse solcher mehrjähriger Kulturen auf Unkräuter wurden in vier Ansätzen untersucht : 1) Vegetationsaufnahmen auf 632 Feldern in Westfrankreich zeigten, dass die Unkrautzusammensetzung zwischen mehrjährigen Futterkulturen und einjährigen Kulturen stark variiert. Der Vergleich von Feldern vor, während und nach mehrjährigen Futterkulturen legte nahe, dass die Pflanzengemeinschaft während solcher Fruchtfolgen zyklisch variiert. Mehrere problematische Unkrautarten wurden während und nach den mehrjährigen Kulturen zurückgedrängt. Das Auftauchen anderer Arten führte jedoch zu einer gleichbleibenden oder leicht erhöhten Pflanzenvielfalt. 2) Ein dreijähriger Feldversuch mit verschiedenen Bearbeitungsoptionen ermöglichte es, die zugrunde liegenden Mechanismen zu untersuchen: Die fehlende Bodenbearbeitung hat das Auflaufen der Unkräuter reduziert und das Überleben der adulten Pflanzen erhöht. Die permanente Vegetationsbedeckung und die häufigen Heuschnitte haben das Wachstum, das Überleben und die Samenproduktion vermindert. 3) Gewächshausexperimente zum Nachwachsen von Unkrautpflanzen nach Heuschnitten zeigten große Unterschiede zwischen verschiedenen Arten und funktionellen Gruppen. Ein Experiment mit zwei Faktoren lässt vermuten, dass die negativen Effekte der Schnitte und der Konkurrenz auf das Unkrautwachstum sich addieren. 4) Spezielle Messungen der Prädation von Unkrautsamen auf den untersuchten Feldern zeigten positive Korrelationen mit der Vegetationsbedeckung, was auf eine besondere Wichtigkeit dieser Ökosystemdienstleistung in ausdauernden Kulturen hindeutet. Die Präferenz von bestimmten Samenarten deutet darauf hin, dass Samenprädation ein weiterer Grund für die beobachteten Änderungen der Unkrautgemeinschaften sein kann.

Les systèmes de culture en Protection Intégrée de Cultures contre les adventices (systèmes PIC-adventices) sont d’un grand intérêt pour réduire significativement l’utilisation des herbicides en les substituant par une combinaison complexe de différentes techniques culturales. L’objectif principal de notre travail a été de contribuer au volet environnemental de l’évaluation multicritère des systèmes PIC-adventices mis en place depuis 2000 sur le domaine expérimental de l’INRA à Bretenière (21). La variabilité des propriétés physiques et hydrodynamiques qui interviennent dans le transfert de l’eau dans le sol a été étudiée dans les 5 systèmes de culture présents sur le site (1 système de référence et 4 systèmes PIC-adventices). Le premier volet du travail a consisté à étudier la rétention de l’eau dans la couche du sol travaillée dans les 5 systèmes de culture. La rétention de l’eau a été comparée dans des échantillons de sol dont la structure a été soit « conservée », soit « remaniée » afin d’identifier la part de la variabilité liée aux caractéristiques physiques intrinsèques du sol de celle induite par les pratiques culturales. Les résultats ont montré que le sol du système de culture avec suppression totale des herbicides, associant une teneur élevée en argile, une teneur moyenne en C organique et un travail superficiel intensif, présentait les valeurs les plus élevées de rétention de l’eau dans la couche de sol travaillée. Le second volet a porté sur l’étude de la variabilité de la conductivité hydraulique proche de la saturation, K(h), aux échelles de l’année culturale et du profil du sol. Pour cette étude, 3 systèmes PIC-adventices (système de culture en semis direct et systèmes de culture avec travail du sol modéré à intensif) ont été caractérisés pendant 2 ou 3 années culturales consécutives, en considérant 3 profondeurs de sol incluant la couche du sol travaillée et l’horizon sous-jacent. Les résultats n’ont montré aucune variabilité de K(h) à l’échelle verticale. Par contre, ils ont mis en évidence les effets à long terme des systèmes PIC-adventices sur les valeurs de K(h). Enfin, le dernier volet a étudié la dynamique de l’eau dans le sol de 2 systèmes PIC-adventices contrastés (système avec conduite du sol en semis direct et système sans herbicide avec travail du sol intensif) en suivant une approche d’hydraulique agricole et par le biais de la caractérisation in situ du potentiel matriciel de l’eau dans le sol et de la mesure des quantités d’eau collectée au moyen de lysimètres à mèche. L’approche utilisée a permis la description de la dynamique de l’eau du sol pendant la période de drainage, la validation du fonctionnement des lysimètres à mèche et le calcul de la surface de sol équivalente qui contribuait au transfert de l’eau dans ces dispositifs. Ces premiers résultats prometteurs nécessiteront néanmoins d’être validés au cours de prochaines campagnes de mesure. L’ensemble des résultats recueillis constituent une première évaluation du fonctionnement physique et hydrodynamique du sol des systèmes PIC-adventices présents sur le dispositif expérimental. La base de données constituée permettra ultérieurement la poursuite du travail engagé en modélisant le transfert de l’eau, de solutés, de pesticides ou d’autres polluants dans le sol de ces systèmes PIC-adventices. / Integrated weed management (IWM)-based cropping systems are of great interest due to their significant reduction of herbicide inputs by using a complex combination of agricultural techniques. The main objective of our work was to contribute to the environmental dimension of the multi-criteria evaluation of IWM-based cropping systems which were set up since 2000 on the experimental domain of the National Institute for Agronomical Research (INRA) to Bretenière (France). The variability of physical and hydrodynamic soil properties involved in the transfer of water through the vadose zone was studied for the 5 cropping systems present on the experimental site (1 reference cropping system and 4 IWM-based cropping systems). The first section of our study consisted in studying the water retention of the tilled soil layers for the 5 cropping systems. Water retention values were compared between undisturbed and repacked soil samples in order to distinguish the part of the variability due to the intrinsic soil properties from the one induced by the agricultural practices. Our results show that the soil from cropping system without any herbicides, associating high clay content, average organic C content and intensive superficial tillage, may increase soil water retention in the tilled soil layer. The second section carried on the study of the annual and vertical variability of the near-saturated hydraulic conductivity, K(h). For this study, 3 IWM-based cropping systems (no-tilled cropping system and moderate to intensive tilled cropping systems) were characterized during 2 or 3 consecutive years and considering 3 depths including the soil tilled layer and the underlying no-tilled layer. The results showed no vertical variability of K(h). However, they highlighted the long-term effect of IWM-based cropping systems on K(h). Finally, the third section studied the soil water dynamics of 2 contrasted IWM-based cropping systems (no-tilled and zero herbicide cropping systems with intensive tillage) by using ,a hydraulic agricultural approach and by the means of the in situ characterization of soil matric potential and the measure of the amounts of water collected with wick lysimeters. This approach allowed the description of the water dynamics during the drainage season, the validation of the wick lysimeters functioning and the calculation of the surface contributing to water flows collected by the wick lysimeters. These promising results will need to be validated during forthcoming campaigns of measurement. All data collected provided constitute a first evaluation of the soil physical and hydraulic functioning of IWM-based cropping systems present on the experimental site. The established database will allow later the pursuit of work by modeling the water transfer and the transport of solutes, pesticides or other pollutants through the vadose zone of these IWM-based cropping systems.