Several aspects of plant-pollinator interactions are presented in the thesis. It contains a review on the open questions of plant-pollinator interactions from single species to complex networks. The following sections document novel results. Firstly, the conservation of complex pollination networks is addressed through the hierarchy of species' importance. In addition, the habitat requirements and interactions of a threatened rare pollinator species are explored. In the following chapters, the results from manipulative approaches applied in the field to plant-pollinator interactions are presented. The effect of pollinator's population decline on pollinators' foraging for pollen is investigated. Moreover, the way plant species loss impact several aspects of pollinator visitation is presented. Lastly, the impact of species removal on plant-pollinator network topology and on species ability of establishing new interactions is investigated.

Tropical habitats face a diverse range of anthropogenic threats. Two common and important threats to tropical biodiversity are invasive species and roads. Invasive plants are proposed to be a major threat to biodiversity worldwide, yet not much is known about their impacts on higher trophic levels, such as insects. Roads and other linear intrusions, such as power lines and railway tracks, are another common aspect of human disturbance in natural landscapes, including tropical forests, and are often linked to the spread of invasive plants. I studied impacts of these two important proximate drivers of habitat disturbance, namely invasive plant species and roads, on habitat use by butterflies in a tropical moist deciduous forest in Western Ghats of India. Invasive plants and roads are expected to modify micro-habitat structure, resources and other aspects of ecology of butterflies and thereby influence how they use space (i.e., micro-habitats within the larger habitat). Because systematic ecological information on tropical butterflies is comparatively limited, I adopted a multi-species approach. I examined space use responses of butterflies to a gradient of lantana cover in the forest and to a road passing through forest. The abundance of different species of butterflies in different micro-habitats was taken as a measure of habitat use. Data was collected over two seasons and at two spatial scales. The two habitat disturbances were found to influence local habitat use by butterflies. But interestingly, species appeared to respond differently, with some showing positive, others negative and some no clear association with road verge or lantana gradient. I then examined whether this variation in response could be understood in terms of species-specific functional traits. Correlating the responses of species to a habitat disturbance with functional traits may provide a way of arriving at general patterns and increase the ability of studies to predict responses. Species with similar trait values are expected to respond similarly to a habitat change driver. I measured morphological traits in 254 butterfly species from India and classified them according to their habitat preferences (based on expert opinion). I first examined relationships between morphological traits, habitat preferences and evolutionary relatedness. I then examined patterns of correlation between these traits and responses to the two habitat disturbances and found that certain traits can help predict responses. Overall, my study suggests that butterfly space use is influenced by roads and lantana, but the response varies across species. These changes in habitat use might have important population or community-level consequences, such as population declines and shifts in community structure and composition; these need to be further examined.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / From december 2006 to april 2008 was conducted a experiment at Universidade Federal de Santa Maria, Santa Maria, RS, Brasil, with objective to verify spatial and temporal distribution of soybean stink bug, with georeferenced surveying. The samplings was accomplished in area of 6,64 hectares of soybean (Glycine max (L.) Merrill), it was divided in a sample grid of the 105 points, spaced in 25 x 25 m. Colonization of the field occurred on soybean s vegetative period, from its adjacent borders to the fields of early soybean and sunflower, by the adults entry from these crops to the experimental area, because later sowing. From the beginning seed until the pod filling of the soybean, the tendency of stink bug population increase by the borders was kept. In the crop 2007/08, due to the sowing have been carried through at that recommended period, the spatial distribution of stink bugs started simultaneously in several points in the field and it did not present clear influence of the adults entry from other areas. In both the crops the populations increase occurred from beginning pod and beginning seed and to species abundance and its population s peaks varied of a crop for another. During overwintering it verified presence of stink bug in the area, feeding itself in the straw on the soil and living on grains fallen into the harvest, or in alternative host plants as Raphanus sp. With objective to study nymphs and adults of Piezodorus guildinii (Westwood, 1837) (Hemiptera: Pentatomidae) behavior on day time on soybean, was conducted a experiment at Universidade Federal de Santa Maria, from april to may 2007. The experimental design employed was a factorial of order fourth, in randomized complete block design (24x3x3x3) (day time x distribution x location x behavior), with five blocks. The stink bugs were observed hourly, during 24 times. The stink bugs plant distribution (top, middle and botton), their location in plant organs (pod, leaf ou HRPR stem, branch, stick or raceme) and their behavior (resting, locomotion, feeding ativity) was registered. During the first and second instars, the nymphs had presented gregarious habit, and a small percentage in movement around of the posture. The nymphs started the feeding activity on second instar, and the locomotion was just around of the posture and the same organ where they were. From the third instar nymphs became more active, with bigger ditribution on plant and feeding more frequently. Locomotion times were near and precedents to feeding times and precedent the dusk, indicating the search for preferential locations for feeding or shelter. Nimphs and adults P. guildinii located preferentially in pod, them whats are associated to repose and feeding activity behaviors. In the nocturnal period, with air temperature lower than in the diurnal period, P. guildinii s nymphs had remained in stems circled by pods, because of the search by sheltered locations. Percentage P. guildinii in feeding activity increased of 7,37% of second-fourth instar for 16,92% in the fifth instar, and 34,77% adult phase. P. guildinii remained, preferentially, in middle plants stracta along the day time, not presenting a clear behavior of distribution in superior plants stracta in defined chedules. Thus, should not be indicated groups ideal schedule's for insecticides application for your control. / De dezembro de 2006 a abril de 2008 foi conduzido um experimento na Universidade Federal de Santa Maria, Santa Maria, RS, Brasil, com o objetivo de verificar a distribuição espacial e temporal de percevejos da soja, através de levantamento georreferenciado. As amostragens foram realizadas em uma área de 6,64 hectares de soja (Glycine max (L.) Merrill) dividida em uma grade amostral com 105 pontos, espaçados em 25 x 25 m. A colonização da lavoura ocorreu no período vegetativo da soja, a partir das bordas adjacentes a lavouras de soja precoce e girassol, pela entrada de percevejos adultos destes cultivos do entorno para a área georreferenciada, em função da semeadura tardia. Desde o início do enchimento de grãos até o enchimento pleno de grãos, a tendência de aumento populacional de percevejos pelas bordas foi mantida. Na safra 2007/08, devido a semeadura ter sido realizada na época recomendada, a distribuição espacial dos percevejos começou simultaneamente em vários pontos na lavoura e não apresentou clara influência da entrada de adultos de outras áreas. Em ambas as safras o aumento das populações ocorreu a partir da formação de legumes e do início do enchimento de grãos e a abundância de espécies e seus picos populacionais variaram de uma safra para outra. Durante todo o período de entressafra, verificou-se a presença de percevejos na área, abrigados na palhada sobre o solo e alimentando-se de grãos caídos na colheita, ou em plantas hospedeiras alternativas como Raphanus sp. Com o objetivo de estudar o comportamento de ninfas e adultos de Piezodorus guildinii (Westwood, 1837) (Hemiptera: Pentatomidae) ao longo do dia em soja, foi conduzido um experimento na Universidade Federal de Santa Maria, de abril a maio de 2007. O delineamento experimental foi um fatorial de quarta ordem em blocos ao acaso (24x3x3x3) (horários x distribuição x localização x comportamentos), com cinco repetições. Os percevejos foram observados de hora em hora, por 24 horas. Registrou-se a sua distribuição nas plantas (terço superior, médio e inferior), a sua localização nos órgãos (legumes, folhas e HRPR haste, ramo, pecíolo ou racemo) e o seu comportamento (repouso, movimento e atividade alimentar). Durante o primeiro e segundo ínstares, as ninfas apresentaram hábito gregário, e uma pequena percentagem de ninfas em movimento em torno da postura. As ninfas começaram a alimentar-se no segundo ínstar, e tornaram-se mais ativas a partir do terceiro ínstar, movimentando-se não só a pequenas distâncias entre órgãos próximos, mas distribuindo-se pelos terços da planta, e alimentando-se mais regularmente. Os horários de movimento foram próximos e antecedentes aos horários de atividade alimentar, e do anoitecer, indicando a procura por locais para alimentação ou abrigo. Ninfas e adultos de P. guildinii localizaram-se preferencialmente em legumes, os quais estão associados aos comportamentos de repouso e atividade alimentar. No período noturno, com a temperatura do ar mais baixa que no período diurno, ninfas de P. guildinii se localizaram em hastes circundadas por legumes, em função da busca por locais abrigados. A percentagem de P. guildinii em atividade alimentar aumentou de 7,37% do segundo ao quarto ínstar para 16,92% no quinto ínstar, e 34,77% na fase adulta. P. guildinii permaneceu, preferencialmente, no terço médio das plantas ao longo do dia, não apresentando um comportamento claro de distribuição no terço superior das plantas em intervalos de horários definidos. Desta forma, não devem ser indicadas faixas de horário ideais para a aplicação de inseticidas para seu controle.