Métodos de aprendizagem automática: um estudo baseado na avaliação e previsão de clientes bancários

Data Mining surge, hoje em dia, como uma ferramenta importante e crucial para o sucesso de um negócio. O considerável volume de dados que atualmente se encontra disponível, por si só, não traz valor acrescentado. No entanto, as ferramentas de Data Mining, capazes de transformar dados e mais dados em conhecimento, vêm colmatar esta lacuna, constituindo, assim, um trunfo que ninguém quer perder. O presente trabalho foca-se na utilização das técnicas de Data Mining no âmbito da atividade bancária, mais concretamente na sua atividade de telemarketing. Neste trabalho são aplicados catorze algoritmos a uma base de dados proveniente do call center de um banco português, resultante de uma campanha para a angariação de clientes para depósitos a prazo com taxas de juro favoráveis. Os catorze algoritmos aplicados no caso prático deste projeto podem ser agrupados em sete grupos: Árvores de Decisão, Redes Neuronais, Support Vector Machine, Voted Perceptron, métodos Ensemble, aprendizagem Bayesiana e Regressões. De forma a beneficiar, ainda mais, do que a área de Data Mining tem para oferecer, este trabalho incide ainda sobre o redimensionamento da base de dados em questão, através da aplicação de duas estratégias de seleção de atributos: Best First e Genetic Search. Um dos objetivos deste trabalho prende-se com a comparação dos resultados obtidos com os resultados presentes no estudo dos autores Sérgio Moro, Raul Laureano e Paulo Cortez (Sérgio Moro, Laureano, & Cortez, 2011). Adicionalmente, pretende-se identificar as variáveis mais relevantes aquando da identificação do potencial cliente deste produto financeiro. Como principais conclusões, depreende-se que os resultados obtidos são comparáveis com os resultados publicados pelos autores mencionados, sendo os mesmos de qualidade e consistentes. O algoritmo Bagging é o que apresenta melhores resultados e a variável referente à duração da chamada telefónica é a que mais influencia o sucesso de campanhas similares.

Characterization and Pathogenicity of Vibrio cholerae and Vibrio vulnificus from Marine environments

The genus Vibrioof the family Vibrionaceae are Gram negative, oxidasepositive, rod- or curved- rodshaped facultative anaerobes, widespread in marine and estuarine environments. Vibrio species are opportunistic human pathogens responsible for diarrhoeal disease, gastroenteritis, septicaemia and wound infections and are also pathogens of aquatic organisms, causing infections to crustaceans, bivalves and fishes. In the present study, marine environmental samples like seafood and water and sediment samples from aquafarms and mangroves were screened for the presence of Vibrio species. Of the134 isolates obtained from the various samples, 45 were segregated to the genus Vibrio on the basis of phenotypic characterization.like Gram staining, oxidase test, MoF test and salinity tolerance. Partial 16S rDNA sequence analysis was utilized for species level identification of the isolates and the strains were identified as V. cholerae(N=21), V. vulnificus(N=18), V. parahaemolyticus(N=3), V. alginolyticus (N=2) and V. azureus (N=1). The genetic relatedness and variations among the 45 Vibrio isolates were elucidated based on 16S rDNA sequences. Phenotypic characterization of the isolates was based on their response to 12 biochemical tests namely Voges-Proskauers’s (VP test), arginine dihydrolase , tolerance to 3% NaCl test, ONPG test that detects β-galactosidase activity, and tests for utilization of citrate, ornithine, mannitol, arabinose, sucrose, glucose, salicin and cellobiose. The isolates exhibited diverse biochemical patterns, some specific for the species and others indicative of their environmental source.Antibiogram for the isolates was determined subsequent to testing their susceptibility to 12 antibiotics by the disc diffusion method. Varying degrees of resistance to gentamycin (2.22%), ampicillin(62.22%), nalidixic acid (4.44%), vancomycin (86.66), cefixime (17.77%), rifampicin (20%), tetracycline (42.22%) and chloramphenicol (2.22%) was exhibited. All the isolates were susceptible to streptomycin, co-trimoxazole, trimethoprim and azithromycin. Isolates from all the three marine environments exhibited multiple antibiotic resistance, with high MAR index value. The molecular typing methods such as ERIC PCR and BOX PCR revealed intraspecies relatedness and genetic heterogeneity within the environmental isolatesof V. cholerae and V. vulnificus. The 21 strains of V. choleraewere serogroupedas non O1/ non O139 by screening for the presence O1rfb and O139 rfb marker genes by PCR. The virulence/virulence associated genes namely ctxA, ctxB, ace, VPI, hlyA, ompU, rtxA, toxR, zot, nagst, tcpA, nin and nanwere screened in V. cholerae and V. vulnificusstrains.The V. vulnificusstrains were also screened for three species specific genes viz., cps, vvhand viu. In V. cholerae strains, the virulence associated genes like VPI, hlyA, rtxA, ompU and toxR were confirmed by PCR. All the isolates, except for strain BTOS6, harbored at least one or a combination of the tested genes and V. choleraestrain BTPR5 isolated from prawn hosted the highest number of virulence associated genes. Among the V. vulnificusstrains, only 3 virulence genes, VPI, toxR and cps, were confirmed out of the 16 tested and only 7 of the isolates had these genes in one or more combinations. Strain BTPS6 from aquafarm and strain BTVE4 from mangrove samples yielded positive amplification for the three genes. The toxRgene from 9 strains of V. choleraeand 3 strains of V. vulnificus were cloned and sequenced for phylogenetic analysis based on nucleotide and the amino acid sequences. Multiple sequence alignment of the nucleotide sequences and amino acid sequences of the environmental strains of V. choleraerevealed that the toxRgene in the environmental strains are 100% homologous to themselves and to the V. choleraetoxR gene sequence available in the Genbank database. The 3 strains of V. vulnificus displayed high nucleotide and amino acid sequence similarity among themselves and to the sequences of V. cholerae and V. harveyi obtained from the GenBank database, but exhibited only 72% homology to the sequences of its close relative V. vulnificus. Structure prediction of the ToxR protein of Vibrio cholerae strain BTMA5 was by PHYRE2 software. The deduced amino acid sequence showed maximum resemblance with the structure of DNA-binding domain of response regulator2 from Escherichia coli k-12 Template based homology modelling in PHYRE2 successfully modelled the predicted protein and its secondary structure based on protein data bank (PDB) template c3zq7A. The pathogenicity studies were performed using the nematode Caenorhabditiselegansas a model system. The assessment of pathogenicity of environmental strain of V. choleraewas conducted with E. coli strain OP50 as the food source in control plates, environmental V. cholerae strain BTOS6, negative for all tested virulence genes, to check for the suitability of Vibrio sp. as a food source for the nematode;V. cholerae Co 366 ElTor, a clinical pathogenic strain and V. cholerae strain BTPR5 from seafood (Prawn) and positive for the tested virulence genes like VPI, hlyA, ompU,rtxA and toxR. It was found that V. cholerae strain BTOS6 could serve as a food source in place of E. coli strain OP50 but behavioral aberrations like sluggish movement and lawn avoidance and morphological abnormalities like pharyngeal and intestinal distensions and bagging were exhibited by the worms fed on V. cholerae Co 366 ElTor strain and environmental BTPR5 indicating their pathogenicity to the nematode. Assessment of pathogenicity of the environmental strains of V. vulnificus was performed with V. vulnificus strain BTPS6 which tested positive for 3 virulence genes, namely, cps, toxRand VPI, and V. vulnificus strain BTMM7 that did not possess any of the tested virulence genes. A reduction was observed in the life span of worms fed on environmental strain of V. vulnificusBTMM7 rather than on the ordinary laboratory food source, E. coli OP50. Behavioral abnormalities like sluggish movement, lawn avoidance and bagging were also observed in the worms fed with strain BTPS6, but the pharynx and the intestine were intact. The presence of multi drug resistant environmental Vibrio strainsthat constitute a major reservoir of diverse virulence genes are to be dealt with caution as they play a decisive role in pathogenicity and horizontal gene transfer in the marine environments.

Robust multimodal face and fingerprint fusion in the presence of spoofing attacks

Anti-spoofing is attracting growing interest in biometrics, considering the variety of fake materials and new means to attack biometric recognition systems. New unseen materials continuously challenge state-of-the-art spoofing detectors, suggesting for additional systematic approaches to target anti-spoofing. By incorporating liveness scores into the biometric fusion process, recognition accuracy can be enhanced, but traditional sum-rule based fusion algorithms are known to be highly sensitive to single spoofed instances. This paper investigates 1-median filtering as a spoofing-resistant generalised alternative to the sum-rule targeting the problem of partial multibiometric spoofing where m out of n biometric sources to be combined are attacked. Augmenting previous work, this paper investigates the dynamic detection and rejection of livenessrecognition pair outliers for spoofed samples in true multi-modal configuration with its inherent challenge of normalisation. As a further contribution, bootstrap aggregating (bagging) classifiers for fingerprint spoof-detection algorithm is presented. Experiments on the latest face video databases (Idiap Replay- Attack Database and CASIA Face Anti-Spoofing Database), and fingerprint spoofing database (Fingerprint Liveness Detection Competition 2013) illustrate the efficiency of proposed techniques.

Machine Learning Algorithms in Heavy Process Manufacturing

In a global economy, manufacturers mainly compete with cost efficiency of production, as the price of raw materials are similar worldwide. Heavy industry has two big issues to deal with. On the one hand there is lots of data which needs to be analyzed in an effective manner, and on the other hand making big improvements via investments in cooperate structure or new machinery is neither economically nor physically viable. Machine learning offers a promising way for manufacturers to address both these problems as they are in an excellent position to employ learning techniques with their massive resource of historical production data. However, choosing modelling a strategy in this setting is far from trivial and this is the objective of this article. The article investigates characteristics of the most popular classifiers used in industry today. Support Vector Machines, Multilayer Perceptron, Decision Trees, Random Forests, and the meta-algorithms Bagging and Boosting are mainly investigated in this work. Lessons from real-world implementations of these learners are also provided together with future directions when different learners are expected to perform well. The importance of feature selection and relevant selection methods in an industrial setting are further investigated. Performance metrics have also been discussed for the sake of completion.

The Pollination Biology and Mating System of a Peripheral Population of Witheringia solanacea (Solanaceae)

Pollinator visitation rates over the life of a flower are determined by pollinator abundance and floral longevity. If flowers are not visited frequently enough, pollen limitation may occur, favoring the evolution of self-compatibility (SC). In plant species with varying SC levels, central populations often are self-incompatible (SI) and peripheral populations are SC. Witheringia solanacea (Solanaceae) is a species that follows this trend with the exception of one population in the Monteverde Cloud Forest Reserve, which is peripheral yet SI. I investigated this population using multiple techniques including floral bagging, pollinator observations, microsatellite analysis, and floral longevity manipulations. My results confirmed the self-incompatibility of the Monteverde population and indicated low but perhaps adequate rates of pollinator visitation per flower per hour. I found reduced genetic diversity at Monteverde and gene flow occurring unidirectionally from San Luis (a central population) to Monteverde. In the greenhouse, there was more of an effect of male than female function on floral longevity, but the largest differences were environmental. Flowers stayed open substantially longer when cool, cloudy weather was simulated and shorter when conditions were hot and sunny. The results indicate that the Monteverde population of W. solanacea is SI because 1) it is unable to maximize its fitness due to gene flow from San Luis and its relatively recent colonization of the area and 2) pollen limitation may not be severe because of supplemental pollinator availability from other Witheringia species in the area and increased floral longevities due to cool and cloudy conditions.

Aplicação de sistemas multi-classificadores no diagnóstico de falhas em motores de indução trifásicos

Equipment maintenance is the major cost factor in industrial plants, it is very important the development of fault predict techniques. Three-phase induction motors are key electrical equipments used in industrial applications mainly because presents low cost and large robustness, however, it isn t protected from other fault types such as shorted winding and broken bars. Several acquisition ways, processing and signal analysis are applied to improve its diagnosis. More efficient techniques use current sensors and its signature analysis. In this dissertation, starting of these sensors, it is to make signal analysis through Park s vector that provides a good visualization capability. Faults data acquisition is an arduous task; in this way, it is developed a methodology for data base construction. Park s transformer is applied into stationary reference for machine modeling of the machine s differential equations solution. Faults detection needs a detailed analysis of variables and its influences that becomes the diagnosis more complex. The tasks of pattern recognition allow that systems are automatically generated, based in patterns and data concepts, in the majority cases undetectable for specialists, helping decision tasks. Classifiers algorithms with diverse learning paradigms: k-Neighborhood, Neural Networks, Decision Trees and Naïves Bayes are used to patterns recognition of machines faults. Multi-classifier systems are used to improve classification errors. It inspected the algorithms homogeneous: Bagging and Boosting and heterogeneous: Vote, Stacking and Stacking C. Results present the effectiveness of constructed model to faults modeling, such as the possibility of using multi-classifiers algorithm on faults classification

Predição de promotores de Bacillus subtilis usando técnicas de aprendizado de máquina

One of the most important goals of bioinformatics is the ability to identify genes in uncharacterized DNA sequences on world wide database. Gene expression on prokaryotes initiates when the RNA-polymerase enzyme interacts with DNA regions called promoters. In these regions are located the main regulatory elements of the transcription process. Despite the improvement of in vitro techniques for molecular biology analysis, characterizing and identifying a great number of promoters on a genome is a complex task. Nevertheless, the main drawback is the absence of a large set of promoters to identify conserved patterns among the species. Hence, a in silico method to predict them on any species is a challenge. Improved promoter prediction methods can be one step towards developing more reliable ab initio gene prediction methods. In this work, we present an empirical comparison of Machine Learning (ML) techniques such as Na¨ýve Bayes, Decision Trees, Support Vector Machines and Neural Networks, Voted Perceptron, PART, k-NN and and ensemble approaches (Bagging and Boosting) to the task of predicting Bacillus subtilis. In order to do so, we first built two data set of promoter and nonpromoter sequences for B. subtilis and a hybrid one. In order to evaluate of ML methods a cross-validation procedure is applied. Good results were obtained with methods of ML like SVM and Naïve Bayes using B. subtilis. However, we have not reached good results on hybrid database

Avaliação de diferentes materiais no ensacamento de pêssegos

O trabalho foi realizado em um pomar comercial de pêssego da cultivar Aurora 2, de três anos de idade, conduzida em sistema de vaso moderno e espaçamento de 6 x 4 m. O delineamento experimental foi em blocos casualizados, utilizando-se de 15 tratamentos, 8 repetições e 16 frutos por tratamento. Os tratamentos foram os seguintes: T1 - Sacolas de TNT branco de 45 gramaturas por m² (45G/m², fechado; T2 - Sacolas de TNT branco de 45 gramaturas por m² (45G/m²) aberto; T3 - Sacolas de TNT branco de 20 gramaturas por m² (20G/m²) fechado; T4 - Sacolas de TNT branco de 20 gramaturas por m² (20G/m²) aberto; T5 - Sacolas de polipropileno microperfurado transparente (furos de 1mm) fechado; T6 - Sacolas de polipropileno microperfurado transparente (furos de 1mm) aberto; T7 - Sacolas de polipropileno microperfurado transparente (furos de 2mm) fechado; T8 - Sacolas de polipropileno microperfurado transparente (furos de 2mm) aberto; T9 - Sacolas de polietileno microperfurado leitoso (furos de 1mm) fechado; T10 Sacolas de polietileno microperfurado leitoso (furos de 1mm) aberto; T11 Sacolas de polietileno microperfurado leitoso (furos de 2mm) fechado; T12 - Sacolas de polietileno microperfurado leitoso (furos de 2mm) aberto; T13 - Sacolas de papel impermeável fechado; T14 - Sacolas de papel impermeável aberto; T15 - Testemunha (sem ensacamento). As sacolas de TNT (Tecido-não-Tecido) e as de papel impermeável tinham as dimensões de 11,5 x 15,0 cm, e as de polipropileno microperfurado, de 13,0 x 20,0 cm. Os resultados obtidos permitiram concluir que as embalagens de polipropileno transparentes podem ser empregadas para o ensacamento de frutos de pessegueiro, uma vez que as mesmas possibilitam a visualização da coloração dos frutos no momento da colheita, não comprometem o desenvolvimento da coloração dos mesmos e apresentam facilidade no manuseio.

Aplicação de técnicas de aprendizado de máquina no reconhecimento de classes estruturais de proteínas

Nowadays, classifying proteins in structural classes, which concerns the inference of patterns in their 3D conformation, is one of the most important open problems in Molecular Biology. The main reason for this is that the function of a protein is intrinsically related to its spatial conformation. However, such conformations are very difficult to be obtained experimentally in laboratory. Thus, this problem has drawn the attention of many researchers in Bioinformatics. Considering the great difference between the number of protein sequences already known and the number of three-dimensional structures determined experimentally, the demand of automated techniques for structural classification of proteins is very high. In this context, computational tools, especially Machine Learning (ML) techniques, have become essential to deal with this problem. In this work, ML techniques are used in the recognition of protein structural classes: Decision Trees, k-Nearest Neighbor, Naive Bayes, Support Vector Machine and Neural Networks. These methods have been chosen because they represent different paradigms of learning and have been widely used in the Bioinfornmatics literature. Aiming to obtain an improvment in the performance of these techniques (individual classifiers), homogeneous (Bagging and Boosting) and heterogeneous (Voting, Stacking and StackingC) multiclassification systems are used. Moreover, since the protein database used in this work presents the problem of imbalanced classes, artificial techniques for class balance (Undersampling Random, Tomek Links, CNN, NCL and OSS) are used to minimize such a problem. In order to evaluate the ML methods, a cross-validation procedure is applied, where the accuracy of the classifiers is measured using the mean of classification error rate, on independent test sets. These means are compared, two by two, by the hypothesis test aiming to evaluate if there is, statistically, a significant difference between them. With respect to the results obtained with the individual classifiers, Support Vector Machine presented the best accuracy. In terms of the multi-classification systems (homogeneous and heterogeneous), they showed, in general, a superior or similar performance when compared to the one achieved by the individual classifiers used - especially Boosting with Decision Tree and the StackingC with Linear Regression as meta classifier. The Voting method, despite of its simplicity, has shown to be adequate for solving the problem presented in this work. The techniques for class balance, on the other hand, have not produced a significant improvement in the global classification error. Nevertheless, the use of such techniques did improve the classification error for the minority class. In this context, the NCL technique has shown to be more appropriated

Controle de pragas do pessegueiro através do ensacamento dos frutos

Objetivou-se, no trabalho, buscar opções ao ensacamento de pêssegos, visando o controle eficiente das principais pragas do pessegueiro. Para isso, instalou-se um experimento em um pomar comercial de pêssego do cultivar Aurora 2, conduzido em sistema de vaso e espaçamento de 6 x 4 m. Os tratamentos foram os seguintes: T1 - Sacolas de TNT branco (45 g/m²) fechado, T2 - Sacolas de TNT branco (45 g/m²) aberto, T3 - Sacolas de TNT branco (20 g/m²) fechado; T4 - Sacolas de TNT branco (20 g/m²) aberto; T5 - Sacolas de polipropileno microperfurado transparente (furos de 1mm) fechado; T6 - Sacolas de polipropileno microperfurado transparente (furos de 1mm) aberto; T7 - Sacolas de polipropileno microperfurado transparente (furos de 2mm) fechado; T8 - Sacolas de polipropileno microperfurado transparente (furos de 2mm) aberto; T9 - Sacolas de polietileno microperfurado leitoso (furos de 1mm) fechado; T10 - Sacolas de polietileno microperfurado leitoso (furos de 1mm) aberto; T11 - Sacolas de polietileno microperfurado leitoso (furos de 2mm) fechado; T12 - Sacolas de polietileno microperfurado leitoso (furos de 2mm) aberto; T13 - Sacolas de papel impermeável fechado; T14 - Sacolas de papel impermeável aberto; T15 - Testemunha (sem ensacamento). de acordo com os resultados concluiu-se que todas as embalagens foram eficientes no controle de moscas-das-frutas e pássaros, porém não para mariposa oriental. As sacolas com abertura no fundo podem ser utilizadas como opção ao ensacamento de pêssegos.

Contribuição de etapas do beneficiamento na qualidade fisiológica de sementes de dois híbridos de milho

O objetivo do trabalho foi verificar a qualidade fisiológica de sementes durante as após várias etapas do beneficiamento de dois híbridos de milho e a contribuição de cada equipamento na melhoria da qualidade das sementes. Foram coletadas amostras de sementes chatas de dois híbridos, peneira 22 L, na recepção e após as fases de secagem, pré-limpeza, limpeza, separação por espessura, mesa gravitacional e quando as sementes encontravam-se prontas para embalagem, totalizando sete etapas do beneficiamento mais uma testemunha (colheita e debulha manual). Realizou-se o teste de germinação, análise de pureza, primeira contagem, índice de velocidade de germinação, envelhecimento acelerado, teste de tretrazólio, teste de frio, comprimento de raiz primária e parte aérea, matéria seca de plântulas, emergência em solo e deterioração controlada. O delineamento experimental foi inteiramente casualizado, com quatro repetições por tratamento/teste, analisando-se cada híbrido separadamente. O beneficiamento promoveu melhoria na qualidade das sementes, sendo que aquelas obtidas após a mesa gravitacional e as prontas para ensaque, foram no conjunto de observações, as que apresentaram melhor qualidade fisiológica.

Distribuição vertical, danos e controle cultural de Stenoma catenifer Walsingham (Lepidoptera: Elachistidae) em pomar de abacate

Stenoma catenifer Walsingham is the major pest in avocado groves (Persea americana Mill.) in the Neotropical region. Management has been difficult for many reasons related to the reduced knowledge on its bioecology and the avocado growing systems. The goal of this work was to study the vertical distribution of S. catenifer in avocado plants, the gagging effect of infested fruit on its survival, and the losses caused. The experiments were conducted in a commercial grove located at the Sao Tomas de Aquino, State of Minas Gerais, Brazil, during the 2001/2002 and 2002/2003 growing season. S. catenifer was observed especially at the lower and central part of the plant, and the inner canopy was the least attacked in comparison with the north, south, east and west quadrants. These results may be important indicators for the use of am adequate sampling procedure. The highest percentage of attacked fruit had one to four larvae; in some cases up to eight individuals per fruit were found. The losses caused by the borer varied during the agricultural season, with figures close to 5% in a single evaluation, and the harvest loss may reach 27%. The bagging of infested fruit caused mortality of S. catenifer, as observed through the evaluation carried out four days later.

Influência do tipo de ensacamento no controle de pragas do pessegueiro

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Ecologia da polinização do buriti (Mauritia flexuosa L. – Arecaceae) na restinga de Barreirinhas, Maranhão, Brasil

O estudo das palmeiras nativas é importante por seu grande valor econômico e na manutenção das comunidades de várias espécies de vertebrados e invertebrados que se alimentam de seus frutos, sementes e folhas. A eficiência na produção dos frutos das palmeiras está diretamente relacionada com a presença de insetos polinizadores, principalmente besouros, abelhas e moscas. A palmeira Mauritia flexuosa, comumente conhecida como buriti, é a espécie mais abundante do Brasil e é também chamada de “árvore da vida”, por ser 100% utilizável. Este trabalho teve como objetivo contribuir para o conhecimento da ecologia da polinização do buriti em ambiente de restinga, no município de Barreirinhas, Maranhão, Brasil. Para tanto, obteve-se dados sobre fenologia reprodutiva, biologia floral, sistema reprodutivo e visitantes florais. Para o acompanhamento fenológico foram selecionados 25 indivíduos de cada sexo, os quais foram observados de agosto/2009 a outubro/2012. As fenofases de floração e frutificação foram relacionadas com as variáveis climáticas através de correlação de Spearman. O processo de abertura e longevidade floral foi acompanhado durante o pico de floração da espécie, verificando-se a viabilidade polínica, a receptividade estigmática, as regiões emissoras de odor e a ocorrência de termogênese. Para determinar o sistema reprodutivo foram feitos testes de polinização cruzada e apomixia. O transporte de grãos de pólen pelo vento foi observado, por meio de lâminas de vidros untadas com vaselina que permaneceram penduradas próximas às inflorescências pistiladas durante 24 horas. Os visitantes florais foram coletados através do ensacamento de 20 inflorescências de cada sexo, sendo classificados de acordo com a frequência e o comportamento. O buriti apresentou padrão fenológico anual, sincrônico e sazonal, com floração de agosto a novembro e pico de queda dos frutos em setembro, o que corresponde à estação seca, diferindo do observado na Amazônia, onde estes eventos fenológicos ocorreram na estação chuvosa. Esta diferença pode ser justificada pela grande disponibilidade de água na região, o que faz com que o buriti não necessariamente dependa das chuvas para florescer. Este fato foi evidenciado pela correlação significativa negativa das fenofases com a precipitação e com a umidade relativa. A forte incidência solar e a disponibilidade de água no ambiente contribuíram para o sucesso na floração e frutificação do buriti. Além disto, fatores bióticos podem ter exercido influência no comportamento fenológico, cuja estratégia reprodutiva adotada parece ser a sincronização da floração e da frutificação com a atividade dos polinizadores e dispersores. Dessa maneira a espécie garante a sua reprodução em um período ótimo para a germinação de sementes e estabelecimento de plântulas. O sistema reprodutivo do buriti é xenogâmico. O conjunto de características florais, aliado à abundância de pólen e ao forte odor leva a crer que essa palmeira tenha como principal estratégia de polinização a cantarofilia, porém o vento também possui grande importância na polinização. Além de apresentar polinização do tipo misto (ambofilia), as flores do buriti atraíram uma grande variedade de visitantes, cuja riqueza foi maior que a observada na Amazônia.

Spatial and temporal flower presentation in Apiaceae-Apioideae

Alle Doldengewächse (Apiaceae), darunter die größte, weltweit verbreitete Unterfamilie der Apioideen, weisen in ihren Blütenständen sehr einheitliche Merkmale auf. Die ‚Doppeldolden´ werden aus kleinen, weißen oder gelben Blüten gebildet und von vielen unspezialisierten Insekten besucht. Der uniforme Eindruck, der damit erweckt wird, ist unter Umständen ein Grund, dass die zugrundeliegende Morphologie bislang wenig untersucht wurde. Gegenstand der vorliegenden Dissertation ist es daher, die ‚verborgene Diversität´ im Blütenstandsbereich der Apiaceae -Apioideen mit dem Ziel darzustellen, den Einfluss der Architektur der Pflanzen auf die Art der Blütenpräsentation in Raum und Zeit und damit auf das Reproduktionssystem der Art zu ermitteln. Im ersten Kapitel zeigt der Vergleich von neun ausgewählten Arten, dass in den selbstfertilen und unspezifisch bestäubten Pflanzen durch Synchronisation und Rhythmik in der Präsentation von Blüten Fremdbefruchtung gefördert wird. Entweder durchlaufen die Pflanzen dabei nur eine getrennte männliche und weibliche Blühphase (Xanthoselinum alsaticum) oder der moduläre Bau der Pflanzen führt zu einer Folge männlicher und weiblicher Blühphasen (multizyklische Dichogamie). Die Diözie in Trinia glauca kann in diesem Zusammenhang als eine Trennung der Blühphasen auf verschiedengeschlechtliche Individuen gesehen werden. Für die andromonözischen Arten wird gezeigt, dass der Anteil funktional männlicher Blüten mit steigender Doldenordnung nicht einheitlich zu- oder abnimmt. Dadurch fungieren die Pflanzen zu verschiedenen Zeiten und mit unterschiedlicher Stärke eher als Pollenrezeptoren oder Pollendonatoren. Es wird deutlich, dass das ‚uniforme Muster‘ der Apioideen mit Dolden verschiedener Ordnungen, dichogamen Blüten und deren diversen Geschlechtsausbildungen ein komplexes Raum-Zeit-Gefüge zur Optimierung des Reproduktionssystems darstellt. Das zweite Kapitel stellt die Ergebnisse von Manipulationsexperimenten (Handbestäubung, Bestäuberabschirmung, Entfernen von Dolden niedriger Ordnung) an Chaerophyllum bulbosum dar, die zeigen, dass das Raum-Zeit-Gefüge in der Präsentation der Blüten der Art erlaubt, flexibel auf Umwelteinflüsse zu reagieren. Es stellt sich heraus, dass mechanische Beschädigungen kaum Einfluss auf den Andromonöziegrad und prozentualen Fruchtansatz der Individuen nehmen. Grundvoraussetzung der Reaktionsfähigkeit ist wiederum deren modulärer Bau. Dieser erlaubt es den Pflanzen, zusammen mit dem andromonöziebedingten Reservoir an - geschlechtlich flexiblen - männlichen Blüten, in den später angelegten Dolden fehlenden Fruchtansatz der Blüten früh blühender Dolden zu kompensieren. Im dritten Kapitel wird eine vergleichende Merkmalsanalyse an 255 Apioideen-Arten vorgelegt, die alle Verwandtschaftskreise, Wuchsformen und Verbreitungsgebiete der Gruppe repräsentieren. Ziel der Analyse war die Identifizierung von Merkmalssyndromen, die den Zusammenhang zwischen Architektur und Reproduktionssystem verdeutlichen sollten. Interessanterweise sind die einzigen Merkmale, die miteinander einhergehen, Protogynie und die graduelle Abnahme männlicher Blüten mit steigender Doldenordnung. Alle anderen Merkmale variieren unabhängig voneinander und erzeugen auf vielen verschiedenen Wegen immer wieder das gleiche Funktionsmuster, das als ‚breeding syndrome‘ der Apioideae gedeutet werden kann. Die Arbeit leistet einen wichtigen Beitrag zum Verständnis der Blütenstände der Apiaceen und darüber hinaus zu morphologischer Variation in ‚unspezialisierten‘ Reproduktionssystemen. Offensichtlich liegt in den Apioideen der Selektionsdruck auf der Aufrechterhaltung der generalisistischen Bestäubung und überprägt alle morphologisch-phylogenetischen Merkmalsvarianten.