Inferêrencia frequentista e bayesiana para análise populacional da cochonilha Dysmicoccus brevipes em abacaxizeiro: alerta de infestação e níveis de agregação espaço-temporal

Abstract

The pineapple requires cultural treatments for its success, making it extremely necessary to use appropriate management, mainly due to its vulnerability to attack by pest insect, such as pineapple mealybug. The levels of infestation and stability of population distribution using dispersal models are important for understanding seasonal population patterns and the ethology of the mealybugs. The objective of the research was to analyze the population of Dysmicoccus brevipes (Hemiptera: Pseudococcidae) over time in the pineapple crop, using models for classical and frequentist Bayesiana approaches. The surveys were conducted under field conditions in a conventional commercial pineapple orchard cv. Pérola, located in the Microregion of Itapororoca-PB. Young seedlings from local production were used and planted at 100 cm x 40 cm spacing. Fertilization of the experiment area was carried out conventionally, using NPK as fertilizer. Insect pests were managed according to the occurrence of D. brevipes. Inspections were carried out once a month on the plants, in three regions: apical, median and basal. The variables considered in the analyzes were the proportion of plants infested with the mealybug and the proportion of plants with colonies. The significance of the isolated factors “farm” and “time” and/or interactions between them was assessed using the likelihood ratio test (LRT). Time was compared using confidence intervals around the observed values, and we decided whether one model performed better than the other using the CI values. The results show that in 15% of cases, it was possible to observe the death of the plant as a result of the insect attack. According to the temporal analyses, it is possible to infer that there is a significant level of D. brevipes infestation, which highlights the search for strategies that are aligned with agricultural sustainability principles. Analysis of the models suggests that the insect is distributed in a cluster, that is, aggregated. Additionally, the suggested sampling plan is based on the absence/presence of insects. The injuries caused by the insect, at the beginning of colonization, were imperceptible without thorough evaluation. Such evidence offers the possibility of carrying out future aerial surveys and even without human intervention in future sampling programs for D. brevipes in pineapple, using remote sensing in agricultural systems.