Reproductive endpoints are generally not considered in regulatory risk assessments used to inform registration decisions for pesticides, and relatively few studies have examined effects of herbicides on reproduction in non-target plants. In two sets of greenhouse experiments using three wild species (Rudbeckia hirta L., Centaurea cyanus L. and Trifolium pratense L), effects on flowering phenology and inflorescence characteristics were investigated following low, drift-equivalent glyphosate exposure at an early bud stage. Weekly post-spray observations included the number of inflorescences, aborted buds and malformed inflorescences. In the experiment focusing on inflorescence characteristics (C. cyanus and T. pratense only), inflorescences and pollen were collected at five weeks post-spray to measure inflorescence dry weight, count the number of reproductive florets, estimate the amount of pollen per floret, and assess pollen germination in vitro. Flower production was adversely affected in all three species, including delays in flowering, significant increases in the number of aborted buds and malformed inflorescences, an overall reduction in the number of inflorescences produced, as well as a reduction in the duration of individual inflorescence bloom time (R. hirta and T. pratense assessed only). Inflorescence dry weight and in vitro pollen germination were significantly reduced for C. cyanus exposed to glyphosate, but not for T. pratense. However, both species experienced a significant reduction in the number of reproductive florets produced per inflorescence in response to glyphosate exposure. Neither species was observed to have significant reductions in the amount of pollen produced per reproductive floret. These results have important implications for risk assessment, demonstrating that current glyphosate use in Canada and elsewhere could be adversely affecting non-target flowering plants in field margins, as well as other taxa that rely on them, particularly pollinators. Les effets sur la reproduction des plantes ne sont généralement pas pris en ligne de compte dans les évaluations réglementaires sur les risques des pesticides lors de leur homologation. De plus, relativement peu d'études ont examiné les effets des herbicides sur la reproduction des plantes non ciblées. Dans deux séries d'expériences en serres avec trois espèces sauvages (Rudbeckia hirta L., Centaurea cyanus L. et Trifolium pratense L), les effets sur la phénologie florale et les caractéristiques des inflorescences ont été étudiés après une faible exposition au glyphosate équivalente à la dérive durant la pulvérisation lorsque les plantes sont au début des boutons floraux. Des observations hebdomadaires post-pulvérisation ont été effectuées sur le nombre d'inflorescences, de bourgeons avortés et d’inflorescences malformées. Dans l'expérience portant sur les caractéristiques des inflorescences (C. cyanus et T. pratense seulement), les inflorescences et le pollen ont été recueillis cinq semaines après la pulvérisation pour mesurer le poids sec des inflorescences, compter le nombre de fleurons reproducteurs, estimer la quantité de pollen par fleur et évaluer la germination du pollen in vitro. La production de fleurs a été affectée chez les trois espèces, y compris des retards de floraison, des augmentations significatives du nombre de bourgeons avortés et d’inflorescences malformées, une réduction globale du nombre d'inflorescences produites et une diminution de la durée de floraison par inflorescence. (R. hirta et T. pratense évalués seulement). Le poids sec des inflorescences et la germination in vitro du pollen ont été significativement réduits chez C. cyanus exposé au glyphosate, mais pas chez T. pratense. Cependant, les deux espèces ont subi une réduction significative du nombre de fleurons reproducteurs produits par inflorescence en réponse à l'exposition au glyphosate. Aucune des deux espèces n'a montré de réduction significative de la quantité de pollen produit par fleuron. Ces résultats ont des implications importantes pour l'évaluation de risques, démontrant que le glyphosate, tel qu’utilisé présentement au Canada et ailleurs, pourrait nuire aux plantes à fleurs non ciblées retrouvées en bordure de champs, ainsi qu'à d'autres taxons qui en dépendent, particulièrement les pollinisateurs.

Quando pulverizado, parte do agrotóxico pode se perder para o ambiente, principalmente por deriva. Com isso, substâncias como glifosato podem atingir áreas não-alvos, como a flora de áreas protegidas que compõem o mosaico agrícola. Sendo as plantas consideradas bioindicadoras, este capítulo objetivou avaliar as respostas morfoanatômicas de Bowdichia virgilioides KUNTH. (Fabaceae) à deriva de glifosato, em áreas protegidas de imóveis rurais em Rio Sono – TO. Foram realizadas duas coletas de material vegetal (folhas), em duas fazendas, para dez indivíduos na Área de Reserva (AR) e dez na Faixa Adjacente à Lavoura (FAL). As folhas coletadas foram reservadas para a realização das análises visual, anatômica e micromorfométrica, sendo escolhidas aleatoriamente. A análise visual foi realizada com a classificação de cinco folhas, de cada indivíduo, em cinco classes, que variaram de 0 a 100% de lesão. Posteriormente, realizou-se o índice de injúria foliar (IIF) e o índice de severidade (IS) para as duas áreas, em cada coleta. Na análise anatômica fez-se cinco cortes, na região mediana de cinco folíolos, de folhas diferentes. Após, realizou-se técnicas usuais de anatomia e utilizouse fotomicroscópio para a obtenção de imagens. Já para a análise micromorfométrica, escolheuse um corte, para cada uma das cinco lâminas de cada indivíduo, de ambas as áreas em cada coleta. Posteriormente, usou-se o software Anati Quanti para mensurações da espessura de cada tecido foliar, sendo tais dados expressos em médias. Com o software estatístico R, verificou-se as similaridades de cada tecido entre a AR e a FAL, assim como, entre coletas realizadas em uma mesma área. Na avaliação visual foram percebidos sintomas de cloroses e necroses. As plantas da FAL, principalmente na segunda coleta, apresentaram os maiores IIFs, assim como os maiores ISs. Possivelmente, estes índices foram influenciados pelo fato destas plantas estarem na borda da área de manejo agrícola; e, a segunda coleta ter sido exposta ao efeito de deriva por maior período, sujeita também, a ocorrência de outras derivas de mais pulverizações na lavoura. Sintomas como plasmólise celular, hiperplasia, proliferação (tecido de cicatrização) e colapso celular foram encontrados. Provavelmente, o herbicida provocou tais sintomas, já que outros autores, em estudos com simulação de deriva de glifosato, inclusive com B. virgilioides verificaram tais sintomas. Na micromorfometria a espessura do parênquima paliçádico (PP) apresentou diferenças nas plantas da AR e FAL na segunda coleta, e, nas plantas da FAL, em todas as coletas. Na segunda coleta, as diferenças encontradas nos tecidos das plantas de áreas diferentes podem ter sido ocasionadas devido a maior exposição, dos indivíduos da FAL, ao sol, possuindo assim maior espessura. A diferença entre os tecidos, das plantas coletadas entre primeira e segunda coleta na FAL, provavelmente aconteceu devido ao alongamento expressivo das células do PP na segunda coleta. Em outros estudos, doses intermediárias, na simulação de deriva, demonstraram um aumento na espessura do PP, o que provavelmente também ocorreu neste trabalho. Diante destes resultados, percebeu-se que, possivelmente, o glifosato pulverizado na lavoura, pela deriva, tem alcançado as áreas protegidas dos imóveis rurais provocando alterações visuais, anatômicas e morfométricas nas plantas de B. virgilioides, nativa do Cerrado. / When sprayed, part of the pesticide can be lost to the environment, mainly by drift. Thus, substances such as glyphosate can reach non-target areas, such as the flora of protected areas that make up the agricultural mosaic. Since the plants are considered bioindicators, this chapter aimed to evaluate the morphoanatomic responses of Bowdichia virgilioides KUNTH. (Fabaceae) derived from glyphosate, in protected areas of rural properties in Rio Sono - TO. Two collections of plant material (leaves) were carried out in two farms, for ten individuals in the Reserve Area (AR) and ten in the Adjacent Range to Farming (FAL). The collected leaves were reserved for the accomplishment of the visual, anatomical and micromorphometric analyzes, being chosen randomly. The visual analysis was performed with the classification of five leaves, from each individual, into five classes, ranging from 0 to 100% of lesion. Subsequently, the foliar injury index (IIF) and the severity index (IS) were performed for both areas, in each collection. In the anatomical analysis five cuts were made, in the median region of five leaflets, of different leaves. After that, we performed usual anatomy techniques and used a photomicroscope to obtain images. For the micromorphometric analysis, a cut was chosen, for each of the five slides of each individual, of both areas in each collection. Subsequently, the Anati Quanti software was used to measure the thickness of each leaf tissue, such data being expressed as averages. With statistical software R, we verified the similarities of each tissue between RA and FAL, as well as between collections performed in the same area. In the visual evaluation, symptoms of chlorosis and necrosis were observed. The FAL plants, mainly in the second collection, had the highest IIFs, as well as the highest ISs. Possibly, these indices were influenced by the fact that these plants are on the edge of the area of agricultural management; and the second collection was exposed to the drift effect for a longer period, also subject to the occurrence of other drifts of more spraying in the crop. Symptoms such as cellular plasmolysis, hyperplasia, proliferation (healing tissue) and cell collapse were found. The herbicide probably provoked such symptoms, since other authors, in studies with simulation of glyphosate drift, including B. virgilioides, have verified such symptoms. In the micromorphometry, the thickness of the palisade parenchyma (PP) showed differences in the RA and FAL plants in the second collection, and in the FAL plants, in all collections. In the second collection, the differences found in the tissues of plants from different areas may have been due to the greater exposure of the FAL individuals to the sun, thus having a greater thickness. The difference between the tissues of the plants collected between the first and second collection in the FAL probably occurred due to the expressive stretching of the PP cells in the second collection. In other studies, intermediate doses in the drift simulation demonstrated an increase in PP thickness, which probably also occurred in this work. In view of these results, it was possible to observe that the glyphosate sprayed on the crop by drift has reached the protected areas of the rural properties causing visual, anatomical and morphometric changes in the plants of B. virgilioides, native to the Cerrado.