No description available.

No description available.

Herbivory by Western Flower Thrips (Frankliniella occidentalis Pergande) (WFT) and drought-stress due to limited water availability are currently two major factors that can severely impact cotton (Gossypium hirsutum L.) production. This dissertation examines the effects of drought-stress on cotton and host-plant resistance (HPR) to WFT in laboratory conditions, and seeks to identify the physiological and morphological mechanisms that underlie drought-tolerance and HPR. A life-history systems-approach was developed that provides a new level of detail for understanding how environmental variation impacts adult female WFT. The approach was illustrated by investigating the combined effects of cotton genotype, periodic drought-stress, and prey availability on the adult female omnivorous thrips using a factorial design. Three treatment conditions were significantly different, none of which were predicted based on prevailing ecological-hypotheses. At the same time, the approach produced three novel insights about WFT life-history and reproductive strategy. The roles of negative photo-taxis and leaf biomechanical properties were investigated as potential mechanisms that influence WFT foraging-decisions on individual cotyledons. Results showed that WFT foraging-decisions could be considered adaptive, but there was limited support for either of the mechanisms investigated. The physiological responses to drought stress and drought recovery were investigated for three transgenic cotton cultivars and an untransformed wild-type (WT). At peak drought, ABA levels, stomatal area, and stomatal apertures in the transgenic isolone, AtRAV1-1 were 48% lower, 27.7%, and 16.3% smaller than WT. These results suggest that AtRAV1-1 was the most drought-tolerant and support the hypothesis that changes in stomatal morphology may have functionally contributed to drought-tolerance. Lastly, I investigated whether changes in phytohormone concentrations associated with periodic-drought stress in four cotton cultivars (three transgenic and WT) were correlated with WFT feeding, fitness and state-dependent reproductive responses (i.e., the relationship between initial weight and reproduction). Results showed that JA-Ile and JA were positively correlated with state-dependent egg viability and fecundity, respectively, and negatively correlated with total egg viability and fecundity, respectively, supporting the hypothesis that JA and JA-Ile underlie the negative effects on WFT reproduction and the associated shift to state-dependent reproduction.

Monoculture agriculture has developed as a result of the Western agricultural growth model, which emphasizes reduced on-farm labor and maximum yield. As a result soil health, which is reliant on a diversity of soil-dwelling organisms, is compromised, pest problems are intensified, and biodiversity is lost when vast land areas are devoted to simplified vegetation schemes. There has been a tremendous rise in interest in alternative cropping schemes. The traditional practice of intercropping has received renewed interest as the emphasis on agricultural growth shifts from a purely development-based model to one of conservation and enhanced biodiversity. Although intercropping has shown promising results in controlling specialist herbivorous insects, how intercropping works is not known. Theories that explain the underlying mechanism of intercropping success include chemical repellency and physical masking. We tested these two theories by creating a simulated intercropping system in mesocosm cages in a laboratory environment. We tested twenty intercrops that varied in their vegetation type, size, and phylogenetic distance for their ability to repel an insect pest that recently invaded into North America, the swede midge (Contarinia nasturtii), from its host plant, Brassica oleracea. We found that different non-host plant treatments significantly influenced larval abundance, which indicates that C. nasturtii responded to some aspect of the varying plant combinations. We found that phylogenetic distance did not influence larval densities. Additionally, non-host plant height and leaf area of non-host plants did not influence larval densities. We found that vegetation type significantly affected larval densities. Brassica oleracea planted in combination with groundcover non-host plants had the fewest number of larvae, followed by B. oleracea planted in combination with vegetables. The highest number of larvae was found on B. oleracea plants planted in combination with herb non-host plants. Our research did not support a chemical repellency or visual masking theory of intercrop success.

Causes and patterns of invasive plant species establishment and success depend broadly upon their ecology, including habitat suitability and interactions with other plants and animals. Here I examine the traits and distribution of invasive plants in Vermont, using spatial analysis, laboratory and field studies. I used GIS to investigate environmental factors correlated with presence of 19 invasive plant species in Vermont campgrounds. My results support the assumption that human dispersal of invasive plant seed and stock may be more important than natural dispersal of these plant species to new sites. I also investigate in-depth the relationships of invasive herbaceous garlic mustard (Alliaria petiolata) with native tree seedlings and co-occurring herbaceous plants in the greenhouse and Vermont forests, respectively. Shade from > 1 m tall A. petiolata plants may effect root:shoot ratios of neighboring tree seedlings and interact with nutrition quality of sites to affect their growth patterns. Invasive plants' integration into novel environments is also mediated by their interactions with native invertebrate species. A. petiolata is associated with a unique assemblage of aboveground invertebrates compared with neighboring native plants. Observations indicate A. petiolata may also serve as an attractant for ants, bees, and wasps who feed from water and nectar at the base of the flower or silique during its flowering and seeding period. These results collectively inform our understanding of plant invasion patterns and management strategies of A. petiolata in Vermont. Community interactions are probably more important than allelopathy in determining the influence of Alliaria petiolata on native ecosystems.

A thesis submitted to the Faculty of Science under the school of Molecular and Cell Biology in fulfilment for requirements for Doctor of Philosophy Degree. February 2016 / The use of synthetic chemical pesticides has several negative implications for the Agricultural industry, which include the development of resistance to the insecticides, crop contamination and the killing of non-target insects. This has brought many scientists in the field of nematology and entomology to investigate biological control agents which can help solve identified challenges and these biocontrol agents have also included entomopathogenic nematodes. The majority of entomopathogenic nematodes species that have been isolated belong to Heterorhabditids and Steinernematids which act as vectors for insect pathogenic bacteria species belonging to the genera, Photorhabdus and Xenorhabdus, respectively. However, other species of nematodes, one of which includes a strain of Caenorhabditis briggsae, have also been shown to act as a vector for an insect pathogenic strain of Serratia marcescens. Oscheius sp. TEL-2014 EPNs have been observed to act as vectors for insect pathogenic bacteria belonging to the genus Serratia. In this study a novel insect pathogenic Serratia sp. strain TEL was isolated from the gut of infective juveniles belonging to a species of Oscheius sp. TEL-2014. Next generation sequencing of the bacteria was conducted by generating genomic DNA paired-end libraries with the Nextera DNA sample preparation kit (Illumina) and indexed using the Nextera index kit (Illumina). Paired-end (2 × 300 bp) sequencing was performed on a MiSeq Illumina using the MiSeq reagent kit v3 at the Agricultural Research Council Biotechnology Platform. Quality control and adapter trimming was performed and the genome was assembled using SPADES. 19 contigs were generated with an average length of 301767 bp and N50 of 200,110 bp. The genome of the Serratia sp. TEL was found to be 5,000,541 bp in size, with a G+C content of 59.1%, which was similar to that of other Serratia species previously identified. Furthermore, the contigs were annotated using NCBI Prokaryotic Genome Automatic Annotation Pipeline. Features of the annotated genome included protein encoding sequence or genes, rRNA encoding genes, tRNA encoding genes, ncRNA sequences and repeat regions. 4,647 genes were found and 4,495 were protein-coding sequences (CDS). The genome contains 36 pseudo genes, 2 CRISPR arrays, 13 rRNA genes with five operons (5S, 16S, 23S), 88 tRNAs genes, 15 ncRNA sequences and 9 frameshifted genes. Several genes involved in virulence, disease, defense, stress response, cell division, motility and chemotaxis were identified. This genome sequence will allow for the investigation of identified genes and that will be critical in furthering the understanding of the insect pathogenicity of Serratia sp. strain TEL. Furthermore, it will provide additional genomic insights about the insect-nematode interactions and thus help us improve their ability to be used as biological control agents in agricultural industries. Oscheius sp. TEL-2014 was tested for its entomopathogenicity and it was found that this species was able to infect and kill two model insects Galleria mellonella and Tenebrio molitor. This new nematode species brought 100% mortality within 72 h post-exposure in G. mellonella and whereas, within 96 hours in T. molitor. Following morphometrics analysis of Oscheius sp. TEL-2014 it was concluded that this nematode is described as a novel entomopathogenic nematode species based on its morphometrics and 18S rRNA gene sequence originality. Whole genome sequencing of Oscheius sp. TEL-2014 inbred lines (7 and 13) was performed using Illumina Hiseq sequencing system and paired ends library preparation protocol. Sequencing reads assembled on Velvet resulted in generation of 75965 contigs (line 7) and 53190 contigs (line 13). Gene prediction tools showed that proteins involved in gene expression and DNA replication are present in Oscheius sp. TEL-2014. The draft genome of Oscheius nematodes will support the improvement and initiation of further studies intended to help us understand the molecular and metabolic processes in this genus.

Les plantes sont au centre d’une grande diversité d’interactions biotiques entre organismes plus ou moins proches qui les exploitent en tant que ressources. L’objectif de cette thèse a été de comprendre comment les infections fongiques de la plante et la diversité des arbres en forêt modifient les interactions arbres-insectes. Nous avons tout d’abord effectué une méta-analyse pour poser le cadre théorique des effets indirects des infections fongiques sur les insectes herbivores associés aux mêmes plantes hôtes. L'effet de l’infection préalable des plantes par les champignons sur les préférences et performances des insectes s’avère généralement négatif. Cependant, la magnitude de cet effet délétère varie selon le mode de vie du champignon, la guilde trophique de l’insecte et la spatialité des interactions (interactions locales vs distantes). Nous avons ensuite analysé de façon empirique les interactions tripartites entre le châtaignier européen (Castanea sativa) et deux de ses bioagresseurs exotiques: le cynips (Dryocosmus kuriphilus), insecte galligène, et Cryphonectria parasitica, champignon pathogène responsable de la maladie du chancre. L'effet sur les taux d’infestation par le cynips de la composition spécifique en essences forestières des forêts de châtaigniers atteintes de chancre a été également étudié. Afin d'identifier les mécanismes sous-jacents aux effets de la diversité des forêts sur cet insecte invasif, les communautés d'insectes parasitoïdes et de champignons endophytes présents dans les galles ont été décrites. Les taux d’infection par le cynips étaient plus faibles dans les mélanges de châtaignier avec du chêne et du frêne que dans des parcelles de châtaignier monospécifiques ou dans les mélanges avec du pin. La composition des forêts influence aussi la composition des communautés de parasitoïdes associés aux galles du cynips mais pas leur abondance, richesse ou diversité. Les communautés de champignons endophytes des galles, étudiées par des méthodes de séquençage de nouvelle génération, sont indépendantes de la composition forestière. Par contre, celles présentes dans les galles différent fortement de celles des tissus foliaires adjacents. Nous avons ainsi apporté de nouvelles preuves que la diversité des plantes et les champignons pathogènes sont des facteurs clés déterminant les interactions plantes-insectes. Etudier comment les plantes interagissent avec leurs insectes et champignons associés, et les mécanismes sous-jacents à l’effet de la diversité des plantes sur ces interactions, doit permettre de mieux comprendre les relations entre diversité et fonctionnement des écosystèmes et de proposer des applications pour la gestion des bio-agresseurs forestiers natifs et exotiques. / Plants are the playground of a large diversity of biotic interactions between related and unrelated organisms exploiting them as common resources. The aim of this thesis was to understand how plant-insect interactions vary with fungal infection of their host plant and plant diversity. I first performed a meta-analysis to provide a theoretical background for plant-mediated effects of fungal infection on herbivorous insects. Overall, I found a negative plant-mediated effect of fungi on both insect preference and performance. However, this effect varied according to fungus lifestyle, insect feeding guild and spatial location of the interactions (local vs distant). Then I experimentally tested plant-fungus-insect tripartite interactions in the particular case of exotic bio-aggressors of the European chestnut (Castanea sativa): the Asian chestnut Gall Wasp (ACGW, Dryocosmus kuriphilus), and the fungal pathogen Cryphonectria parasitica, the causal agent of chestnut blight. I performed an observational study, in natural chestnut forest stands in Italy, where I tested how ACGW infestation rates vary with the tree species composition. I also investigated the mechanisms underlying plant diversity effects on the invasive pest, with a particular focus on its natural enemies such as insect parasitoids and endophytic fungi. ACGW infestation rates was lower in oak and ash chestnut mixtures compared to monocultures or pine-chestnut mixtures. Plot composition also influenced ACGW parasitoid community composition but not their abundances, diversity or richness. Endophytic communities of galls, described by using next generation sequencing methods, did not vary with plot composition. However, they strongly differed from surrounding leaf tissues. We thus provided evidence that plant diversity and fungal pathogens are key drivers of plant-insect interactions. Understanding how plants interact with associated insects and fungi, and mechanisms underlying plant diversity effect on these interactions, will improve our knowledge on diversity-ecosystem functioning relationships and will have practical applications for the management of native and exotic forest pests.

As plantas reconhecem e produzem substâncias voláteis específicas para a atração de parasitóides após o ataque de um herbívoro, num processo conhecido como defesa indireta. Contudo, a capacidade dessas plantas em processar e liberar novos compostos voláteis sob o ataque de um herbívoro não hospedeiro permanece ainda inexplorado. Em vista disso, o presente trabalho buscou investigar o efeito dos voláteis emitidos por planta de milho sob o ataque de uma praga até então não hospedeira (traça-das-crucíferas, Plutella xylostella) comparado ao daqueles emitidos por planta de milho sob ataque de um hospedeiro tradicional (lagarta-do-cartucho-domilho, Spodoptera frugiperda), sobre esses insetos e seus respectivos parasitóides, Apanteles piceotrichosus e Campoletis flavicincta. Os bioensaios com os parasitóides foram conduzidos em fotofase, sendo oferecidos a estes, voláteis de plantas de milho sadias, voláteis liberados nos intervalos de 0-1, 5-6 e 24-25h após tratamento em fotofase por dano mecânico ou herbivoria simulada, e voláteis liberados no intervalo de 5-6h após esses mesmos tratamentos em escotofase. Já os ensaios com as mariposas foram conduzidos em escotofase, sendo oferecidos a estas, voláteis de plantas de milho sadias, voláteis liberados nos intervalos de 5-6h após tratamento em fotofase por dano mecânico ou herbivoria simulada, e voláteis liberados no intervalo de 0-1 e 24- 25h após esses mesmos tratamentos em escotofase. Adicionalmente, foram testados, para P. xylostella, voláteis de plantas de couve-manteiga sadia, e para A. piceotrichosus, voláteis de couve-manteiga sadia, couve-manteiga danificada mecanicamente e couve-manteiga atacada por lagartas de P. xylostella. As fêmeas de C. flavicincta apresentaram atratividade para voláteis emitidos pelas plantas de milho no intervalo de 5 a 6 horas após tratamento com regurgito de S. frugiperda em escotofase. Curiosamente, não apresentou atratividade pelos voláteis liberados nesse mesmo intervalo de tempo após indução com regurgito em fotofase. As fêmeas acasaladas de S. frugiperda foram atraídas por voláteis de plantas de milho sadia e voláteis liberados nos intervalos de 5-6 e 24-25h após dano mecânico ou tratamento da planta com regurgito deste herbívoro. Porém, preferiu voláteis de plantas sadias aqueles de plantas tratadas com regurgito em fotofase. Esses resultados demonstraram que esses insetos são capazes de discriminar entre misturas de voláteis presentes em seu habitat natural, onde ocorre a relação tritrófica milho (planta hospedeira) S. frugiperda (herbívoro) C. flavicincta (parasitóide). Da mesma forma, na relação couve-manteiga (planta hospedeira) P. xylostella (herbívoro) A. piceotrichosus (parasitóide), as fêmeas de P. xylostella foram atraídas pelos voláteis de couve-manteiga sadia, assim como as fêmeas de A. piceotrichosus foram atraídas pelos voláteis de couve-manteiga atacadas por lagartas deste herbívoro. No caso da simulação de uma nova relação, milho (planta não-hospedeira) P. xylostella (herbívoro) C. flavicincta (parasitóide), tanto fêmeas de A. piceotrichosus, quanto de P. xylostella, não foram capazes de responder aos voláteis de plantas de milho sadias, danificadas mecanicamente e danificadas mecanicamente + regurgito de P. xylostella. A determinação desses mecanismos poderá ser útil para maior compreensão do contexto evolutivo entre plantas e insetos e obtenção de novos avanços no manejo e controle biológico de pragas. / Plants recognize and produce specific volatile substances that attract parasitoids after the herbivore attack, characterizing a process known as indirect defense. However, the ability of these plants in processing and releasing novel volatile compounds elicited by a non-host herbivore attack has been poorly explored. Regarding this, the current study aimed to investigate the effect of volatiles emitted by maize plants under the attack of a pest which is not a host so far, diamondback moth, Plutella xylostella compared to those emitted by a common host, fall armyworm, Spodoptera frugiperda on the behavior response of these insects and their respective parasitoids Apanteles piceotrichosus e Campoletis flavicincta. The bioassays with the parasitoids were conducted during photophase and they were exposed to volatiles from undamaged maize, volatiles released at the time intervals 0-1, 5-6 and 24-25h after the treatment of mechanical damage or simulated herbivory during photophase, and volatiles released at the time interval 5-6h after these same treatments in scotophase. The bioassays with the moths were carried out in scotophase and they were exposed to volatiles from undamaged maize, volatiles released at the time interval 5-6h after the treatment of mechanical damage or simulated herbivory during photophase, and volatiles released at time intervals 0-1 and 24-25h after these same treatments in scotophase. Additionally, for P. xylostella volatiles from undamaged kale were tested while for A. piceotrichosus it was tested volatiles from undamaged kale, mechanically damaged and P. xylostella caterpillar damaged kale. C. flavicincta females were attracted to volatiles emitted by the maize plants at the interval 5-6h after the treatment with the S. frugiperda regurgitate in scotophase. Curiously, they were not attracted to volatiles released at the same time interval after the induction elicited by the regurgitate in the photophase. S. frugiperda mated females were attracted by volatiles from undamaged plants and volatiles released at time intervals 5-6 and 24- 25h after the mechanical damage or treated with the regurgitate of this herbivore. Nevertheless, they preferred the volatiles from undamaged maize to the plants treated with the regurgitate during photophase. These results demonstrated that these insects are able to distinguish among the volatile blends present in their natural habitat where it occurs the tritrophic relationship maize (host plant) S. frugiperda (herbivore) C. flavicincta (parasitoid). In the same way, in the relationship kale (host plant)- P. xylostella (herbivore) A. piceotrichosus (parasitoid), P. xylostella females were attracted by the volatiles of undamaged kale as well as the A. piceotrichosus females were attracted to volatiles emited by caterpillar-damaged kale. In the case of simulating a new relationship, maize (non-host plant) P. xylostella (herbivore) C. flavicincta (parasitoid), A. piceotrichosus females and P. xylostella were not able to respond to undamaged maize, mechanically damaged and mechanically damaged+ P. xylostella regurgitate. The determination of these mechanisms can be useful for a better understanding of the evolution context between plants and insects and for obtaining new advances in the management and biological pest control.

O estudo foi dividido em capítulos que incluíram diversos aspectos da ecologia química de insetos e espécies de Piperaceae como se seguem: O Capítulo 1 descreve as informações taxonômicas, a história natural dos insetos, as observações de campo e a organização das espécies hospedeiras de Piperaceae segundo as preferências alimentares dos insetos. Observou-se forte especificidade química de espécies de Coleoptera (Naupactus bipes) e Homoptera (Membracis foliata, Callocanophora sp., e Aethalium reticulatum) por Piper aduncum, P. gaudichaudianum, P. arboreum e P. hispidum que possuem ácidos benzóicos prenilados como produtos predominantes. As Lepidoptera (Quadrus u-lucida, Heraclydes brasiliensis e H. helicorides) constituíram-se num segundo grupo de insetos para o qual constatou-se forte especificidade. Nesse caso as espécies de Piperaceae, P. regnellii e P. solmsianum, são nitidamente acumuladores de neolignanas e/ou lignanas. Os Capítulos 2 e 3 (coleópteros - besouros e lepidópteros - borboletas, respectivamente) descrevem os resultados relacionados aos estudos das reações que ocorreram durante o processo de digestão de folhas de espécies de Piper, seqüestros de metabólitos secundários pelos insetos. O estudo de possíveis fatores atrativos presentes em óleos essenciais foram investigados mediante o uso de ensaios eletrofisiológicos. Foram observadas reações de desmetilações em lignanas tetraidrofurânicas, esterificação de ácidos benzóicos, hidrólise de amidas, reação do tipo ozonólise de lignanas, neolignanas e fenilpropanóides durante o processo digestivo das borboletas (larvas) e dos besouros adultos. Foram observados seqüestros de neolignanas das folhas e raízes de P. regnellii pelas larvas da borboleta Heraclides hectorides e ácidos benzóicos prenilados das raízes de P. gaudichaudianum pelas larvas do besouro Naupactus bipes. Os experimentos de EAG com antenas do besouro N. bipes indicou que os óleos de P. gaudichaudianum, P. regnellii e P. hispidum foram ativos, e tal comportamento foi confirmado por observações em campo, sendo que as respostas mais intensas foram observadas para as fêmeas em relação às respostas para os machos. A análise de GC/EM-EAD permitiu a determinação dos compostos bioativos como os monoterpenos α- pineno, ß-pineno e ß-mirceno. O Capítulo 4 descreve as espécies de homópteros que têm especificidade por P. gaudichaudianum, P. arboreum e P. aduncum que são acumuladoras de ácidos benzóicos 7 prenilados. Outros possíveis fatores determinantes para tal especificidade foram atribuídos ao baixo teor de metabólitos secundários nas seivas, alto teor de micronutrientes, baixo teor de macronutrientes, ausência de lignanas ou neolignanas nas seivas, lignificação com predominância de resíduo siringila (S). O Capítulo 5 descreve a determinação e caracterização estrutural dos 49 metabólitos secundários envolvidos neste estudo. O Capítulo 6 descreve o estudo da estabilidade de lignanas tetraidrofurânicas realizado em função da rara ocorrência na natureza de isômeros com a configuração toda cis. O estudo da estabilidade para os dez possíveis estereoisômeros para a lignana tetraidrofurânica através do cálculo do funcional de densidade, B3LYP, com a base 6- 316(dp) mostrou que a configuração toda cis é de fato menos estável do que a toda trans. / The study was presented in six chapters that included several aspects of chemical ecology of insects and Piperaceae species as following: The Chapter 1 describes taxonomical aspects, natural history of the insects and field observation and the organization of hosts Piperaceae according to the observed feeding preferences of the associated insects. It was observed strong chemical specificity of Coleoptera and Homoptera species by Piper aduncum, P. gaudichaudianum, P. arboreum and P. hispidum which contain prenylated benzoic acids. The Lepidoptera showed preference for species containing neolignanas and/or lignanas such P. regnellii and P. solmsianum. The Chapters 2 and 3 (coleopterous - beetles and lepdopterous-butterflies, respectively), were addressed for the studies of biotransformation reactions, sequestration and attractive assays. The study of possible factors present attractions in essential oils was investigated by the use of electrophysiological essays. It were observed several reactions as esterification, hydrolyses and ozonolyse-type for lignans, neolignans, phenylpropanoids and amides occurring in from species of Piperaceae during the digestive process of the butterflies and of beetles. It was observed sequestration of neolignans of P. regnellii by larvae of Heraclides hectorides and prenylated of benzoic acids from roots of P. gaudichaudianum by the larva of the beetle Naupactus bipes. The experiments of EAG using antennas of the beetle N. bipes showed that the oils of P. gaudichaudianum, P. regnellii and P. hispidum were active and such results were in agreement with field observations and the most intense response was observed for the females in relation to the response for the males. The analyses of GC/EM-EAD allowed the determination of the bio-actives compounds as the monoterpenes α-pinene, β-pinene and β-mircene. The Chapter 4 describes the homopteros species that have specificity for the P. gaudichaudianum, P. arboreum and P. aduncum accumulative of prenylated benzoic acids. Further determinant factors for specificity included: low percentage of secondary metabolites in saps, high percentage of micro-nutrients and low percentage of macro-nutrients, absence of lignans or neolignans in the saps, lignifications with predominance of siringyl residues (S) which confer a pattern related to Angiosperms. The Chapter 5 describes the determination and structural characterization of the 49 secondary metabólitos involved in this study. The Chapter 6 shows the study of stability versus natural occurrence of tetrahydrofuran lignans was carried out due to the rare occurrence of all-cis configuration. The study of the stability for the ten possible stereoisomers for the tetrahydrofuran lignans was through the calculation of the functional of density, B3LYP, with the base 6-316(dp), it showed that the all-cis configuration is less stable than the all-trans tetrahydrofuran lignans.

A interação mutualística, espécie-especifica, vespas de figo-figueiras envolve dois processos antagonísticos, predação de sementes e polinização, realizadas por vespas da família Agaonidae. Sabe-se que a larva da vespa se alimenta de tecido da semente durante seu desenvolvimento, mas o processo pelo qual isto ocorre é pouco conhecido, não se sabendo até que ponto a larva depende do desenvolvimento da semente. Neste trabalho foi estudada a biologia larval de Pegoscapus tonduzi, polinizadora de Ficus citrifolia (Moraceae). O estudo foi realizado em plantas de F. citrifolia presentes no campus da Universidade de São Paulo em Ribeirão Preto/SP, durante o período de julho de 2007 a agosto de 2008. Para tal, quatro coortes de vespas foram estudas, nas quais cerca de cinco figos foram coletados em intervalos de dois dias, ao longo do ciclo de desenvolvimento larval. Os figos foram dissecados para observação das larvas e para a obtenção de ovários/galhas da planta para o preparo de lâminas histológicas. Os resultados mostraram que o processo de predação de sementes realizado pelas larvas de P. tonduzi é elaborado, com o desenvolvimento larval intimamente relacionado aos processos embriogênicos da planta. O desenvolvimento larval apresenta quatro instares, detectados pela mudança de tamanho e formadas larvas. A duração do ciclo de vida de P. tonduzi foi de 40 a 70 dias, aproximadamente, correlacionando-se negativamente com a temperatura ambiente no período. O ovo é depositado na região próxima à base do estilete, entre o nucelo e o tegumento interno do ovário da planta. Nessa fase, observase o surgimento do embrião vegetal, indicando que o ovário em que larva se desenvolve foi fertilizado. No segundo estádio larval, o inseto migra para a região micropilar e passa a se alimentar oralmente do endosperma da planta. Nesta fase, ainda, nota-se o desaparecimento do embrião vegetal, sugerindo que este é consumido pela larva da vespa. Os resultados sugerem a existência de um ajuste evolutivo fino entre inseto-planta, uma vez que o desenvolvimento da larva da vespa de figo parece depender da fertilização e conseqüente formação do endosperma. / The species-specific mutualistic interaction between fig trees and fig wasps engages two antagonist processes: seed predation and pollination, both achieved by wasps of the Agaonidae family. It is well known that fig wasp larvae feed on seed tissues for their development but the process itself is poorly known. Here is a study on the larval biology of Pegoscapus tonduzi, pollinator of Ficus citrifolia (Moraceae). The study was carried out between July 2007 and August 2008 at the Sao Paulo University campus in Ribeirao Preto, Sao Paulo State. Four cohorts of fig wasps were studied; for each cohort, about five figs were sampled each two days until complete larval cycle. Sampled figs were dissected for larvae observation and measurement, and flower ovaries/galls subsamples destined to histological study. Results showed that seed predation by fig wasps is a complex process, with larval development closely related to plant embryogenesis. Four larval instars were determined by changes on larva size and shape. Larval cycle extended from 40 to 70 days, showing a negative relation with environmental temperature. The egg is laid near the style insertion, between nucleus and inner integument of the flower ovary. Vegetal embryo was observed along with first larval instar, indicating that fertilization took place inside the ovary where larva develops. At second instar, the larva migrates to the micropilar region and begins to feed orally from endosperm. In this phase, the embryo disappears, suggesting that it is consumed by the larva. Our results suggest a fine tune evolutionary insect-plant adjustment, as fig wasp larvae seems to depend on ovary fertilization and endosperm development.