Abstract The quality of the raw material for chocolate manufacture depends on the control of cacao seed fermentation. Practices affect the diversity and metabolic activities of spontaneous yeasts in fermentations. The control of the initial inoculum is important to standardize fermentations and to improve the final seed quality. We conducted fermentations using non-cleaned or cleaned fermentation boxes to evaluate the dynamics of the microorganism populations and their effect in seed composition in two consecutive years. The isolated yeasts were identified by sequencing the D1-domain of the 26S ribosomal genes. Pulp and cotyledon parameters were monitored during fermentations. Wickerhamomyces anomalus was the dominant yeast in the fermentations, but the species was less represented in cleaned boxes. In the cleaned-box fermentation, there was a delay in temperature rise of the fermentation, which retarded the suitable moment for seed-mass turning. The reduction in seed-pulp quality caused by restricted rainfall in one year limited the seed-mass heating and voided the seed turning. Thus, under substrate restriction, the initial reduced inoculum caused by cleaning the boxes impaired fermentation, mainly under environmental conditions that compromise the quality of the pulp. However, this variation had little effect on the desired degradation of cotyledon components, substrates for flavor precursors. noncleaned non years D1domain Ddomain D1 domain D S genes cleanedbox box seedmass mass turning seedpulp pulp Thus restriction However components precursors

A obtenção de soros policlonais contra fitopatógenos tem por finalidade a elaboração de imunoensaios que permitam identificar e caracterizar os mesmos com rapidez e a custos baixos. Soros policlonais foram produzidos contra extratos das hifas de isolados de Crinipellis perniciosa, agente causal da vassoura-de-bruxa, coletados de cacaueiro (Theobroma cacao) nos estados do Pará e Bahia para identificação e caracterização de isolados de diversas regiões e hospedeiros. O soro denominado de 1 foi produzido a partir do isolado ESJOH-1 (originário do Pará), e o soro 2 foi produzido com o isolado CP-85 (originário da Bahia). Não houve reação cruzada dos soros contra extratos protéicos das espécies Alternaria solani, Marasmius sp., Oudemansiella canarii e Verticillium fungicola. Os soros apresentaram reação cruzada de alta afinidade com extratos de Moniliophthora roreri e de baixa a média afinidade com Marasmius cladophyllus. A caracterização dos isolados foi feita com amostras do biotipo-C coletados no Amazonas, Bahia, Mato Grosso, Rondônia e Pará. As reações sorológicas dos dois soros não permitiram diferenciar os isolados do Pará e da Bahia, contra os quais foram produzidos, e também não reagiram diferencialmente com isolados de outros estados. O reconhecimento dos antígenos entre os soros foi variável, pois o soro 1 e soro 2 apresentaram reação sorológica diferenciada para alguns isolados. Os soros produzidos apresentaram alta afinidade contra isolados dos biotipos-C e S, independente de sua origem geográfica, podendo ser usados para identificação.

The production of polyclonal sera against plant pathogens allows the development of immunoassays to quickly identify and characterize the pathogens at low cost. Polyclonal sera were produced against Crinipellis perniciosa, the causal agent of witches' broom disease of cacao (Theobromae cacao) using hyphae extracts from isolates collected in Pará and Bahia states, to identify and distinguish isolates from various regions and hosts. Serum 1 was produced from isolate ESJOH-1 (collected in Pará) and serum 2, from isolate CP-85 (from Bahia). There was no cross reaction between the two sera with mycelia extracts from Alternaria solani, Marasmius sp., Oudemansiella canarii and Verticillium fungicola. The sera presented high affinity cross reaction with Moniliophthora roreri and medium- to low cross-reaction against Marasmius cladophyllus. The sera were evaluated against isolates from biotype-C, collected in the states of Amazonas, Bahia, Mato Grosso, Rondônia and Pará. The serological reaction of both sera did not allow to distinguish the isolates' site of origin (Pará and Bahia), nor did they differentiate isolates of biotype C. The antigen binding sites were different for serum 1 and 2, because both exhibited different response to the same strains. However, these differences in affinity did not allow for differentiating the origin of collection, host or biotype, but allowed for the identification of C. perniciosa belonging to biotypes C and S.