Strawberry growers in Florida suffer from a lack of efficient and accurate yield forecasts for strawberries, which would allow them to allocate optimal labor and equipment, as well as other resources for harvesting, transportation, and marketing. Accurate estimation of the number of strawberry flowers and their distribution in a strawberry field is, therefore, imperative for predicting the coming strawberry yield. Usually, the number of flowers and their distribution are estimated manually, which is time-consuming, labor-intensive, and subjective. In this paper, we develop an automatic strawberry flower detection system for yield prediction with minimal labor and time costs. The system used a small unmanned aerial vehicle (UAV) (DJI Technology Co., Ltd., Shenzhen, China) equipped with an RGB (red, green, blue) camera to capture near-ground images of two varieties (Sensation and Radiance) at two different heights (2 m and 3 m) and built orthoimages of a 402 m² strawberry field. The orthoimages were automatically processed using the Pix4D software and split into sequential pieces for deep learning detection. A faster region-based convolutional neural network (R-CNN), a state-of-the-art deep neural network model, was chosen for the detection and counting of the number of flowers, mature strawberries, and immature strawberries. The mean average precision (mAP) was 0.83 for all detected objects at 2 m heights and 0.72 for all detected objects at 3 m heights. We adopted this model to count strawberry flowers in November and December from 2 m aerial images and compared the results with a manual count. The average deep learning counting accuracy was 84.1% with average occlusion of 13.5%. Using this system could provide accurate counts of strawberry flowers, which can be used to forecast future yields and build distribution maps to help farmers observe the growth cycle of strawberry fields.

Soybean (Glycine max. L. Merrill) production in Paraguay has increased dramatically during the last decade and the country is now the fourth largest soybean exporter in the world, producing about 3% of the world's soybean production. This paper explored associations between soybean yield in eastern Paraguay and the El Niño Southern Oscillation (ENSO) phases. Historical soybean yield data from official government sources were detrended to remove the effects of technological advances, and yield residuals were computed. In addition, differences in mean precipitation among ENSO phases were investigated under the context of crop development phases. The CSM‐CROPGRO soybean model was used to simulate soybean development for two locations representing the most important soybean producing areas in Paraguay. Influences of ENSO phases on mean precipitation during planting and blooming, blooming and seed podding, and from young pods to physiological maturity were explored through tests of differences in the central tendency. Relative yield residuals during El Niño years were positive six out of seven events and varied from − 9.4 to + 24.2% for the 1991/1992 and 2002/2003 cropping seasons, respectively. During La Niña years, calculated residuals were negative for three out of four events and varied from − 37.9 to + 1.5% for the 2005/2006 and 1988/1989 cropping seasons, respectively. Analysis of precipitation records showed significantly lower precipitation levels between planting and blooming during La Niña years than during El Niño years. Differences in mean precipitation during blooming and beginning of seed formation were found to be not significant. Mean precipitation between seed podding and crop maturity was found to be significantly lower during La Niña years than during El Niño years in one of the locations studied. Copyright © 2007 Royal Meteorological Society

Epidemics of sugarcane orange rust (caused by Puccinia kuehnii) in Florida are largely influenced by prevailing weather conditions. In this study, we attempted to model the relationship between weather conditions and rust epidemics as a first step toward development of a decision aid for disease management. For this purpose, rust severity data were collected from 2014 through 2016 at the Everglades Research and Education Center, Belle Glade, Florida, by recording percentage of rust-affected area of the top visible dewlap leaf every 2 weeks from three orange rust susceptible cultivars. Hourly weather data for 10- to 40-day periods prior to each orange rust assessment were evaluated as potential predictors of rust severity under field conditions. Correlation and stepwise regression analyses resulted in the identification of nighttime (8 PM to 8 AM) accumulation of hours with average temperature 20 to 22°C as a key predictor explaining orange rust severity. The five best regression models for a 30-day period prior to disease assessment explained 65.3 to 76.2% of variation of orange rust severity. Prediction accuracy of these models was tested using a case control approach with disease observations collected in 2017 and 2018. Based on receiver operator characteristic curve analysis of these two seasons of test data, a single-variable model with the nighttime temperature predictor mentioned above gave the highest prediction accuracy of disease severity. These models have potential for use in quantitative risk assessment of sugarcane rust epidemics.

O objetivo deste trabalho foi avaliar a redução do vigor vegetativo da cobertura vegetal do Pampa do Brasil e do Uruguai, por meio da identificação de tendências negativas em séries temporais de imagens. Utilizaram-se séries temporais de imagens de NDVI/EVI do sensor Modis, de 2000 a 2011; imagens de índices de umidade do solo do "climate forecast system reanalysis"; e dados de precipitação pluvial de estações meteorológicas. O estudo quantificou tendências lineares e não lineares nas séries de NDVI e EVI, em áreas de campos. Na tendência monotônica de Mann-Kendall, a 5% de probabilidade, 81,9% da área total estudada foi significativa com o NDVI, e 74,8%, com o EVI; no entanto, o EVI apresentou contraste superior na estimativa dos parâmetros. Os resultados mostraram maior sinal negativo a oeste, com valores médios de R²>0,15, r<-0,3 e τ <-0,15 na tendência dos índices de vegetação, e tendência decrescente para NDVI, EVI e precipitação pluvial, com menores valores médios de umidade do solo. A tendência negativa dos índices de vegetação, relacionada à combinação da ocorrência de deficit hídrico em solos rasos com o sobrepastoreio, indica alterações no padrão de cobertura vegetal do Pampa, com redução do vigor vegetativo.