Using Convolutional Neural Networks to Efficiently Extract Immense Phenological Data From Community Science Images

Community science image libraries offer a massive, but largely untapped, source of observational data for phenological research. The iNaturalist platform offers a particularly rich archive, containing more than 49 million verifiable, georeferenced, open access images, encompassing seven continents and over 278,000 species. A critical limitation preventing scientists from taking full advantage of this rich data source is labor. Each image must be manually inspected and categorized by phenophase, which is both time-intensive and costly. Consequently, researchers may only be able to use a subset of the total number of images available in the database. While iNaturalist has the potential to yield enough data for high-resolution and spatially extensive studies, it requires more efficient tools for phenological data extraction. A promising solution is automation of the image annotation process using deep learning. Recent innovations in deep learning have made these open-source tools accessible to a general research audience. However, it is unknown whether deep learning tools can accurately and efficiently annotate phenophases in community science images. Here, we train a convolutional neural network (CNN) to annotate images of Alliaria petiolata into distinct phenophases from iNaturalist and compare the performance of the model with non-expert human annotators. We demonstrate that researchers can successfully employ deep learning techniques to extract phenological information from community science images. A CNN classified two-stage phenology (flowering and non-flowering) with 95.9% accuracy and classified four-stage phenology (vegetative, budding, flowering, and fruiting) with 86.4% accuracy. The overall accuracy of the CNN did not differ from humans (p = 0.383), although performance varied across phenophases. We found that a primary challenge of using deep learning for image annotation was not related to the model itself, but instead in the quality of the community science images. Up to 4% of A. petiolata images in iNaturalist were taken from an improper distance, were physically manipulated, or were digitally altered, which limited both human and machine annotators in accurately classifying phenology. Thus, we provide a list of photography guidelines that could be included in community science platforms to inform community scientists in the best practices for creating images that facilitate phenological analysis.

Effectiveness of Common Preprocessing Methods of Time Series for Monitoring Crop Distribution in Kenya

Accurate crop identification and spatial distribution mapping are important for crop production estimation and famine early warning, especially for food-deficit African agricultural countries. By evaluating existing preprocessing methods for classification using satellite image time series (SITS) in Kenya, this study aimed to provide a low-cost method for cultivated land monitoring in sub-Saharan Africa that lacks financial support. SITS were composed of a set of MODIS Vegetation Indices (MOD13Q1) in 2018, and the classification method included the Support Vector Machine (SVM) and Random Forest (RF) classifier. Eight datasets obtained at three levels of preprocessing from MOD13Q1 were used in the classification: (1) raw SITS of vegetation indices (R-NDVI, R-EVI, and R-NDVI + R-EVI); (2) smoothed SITS of vegetation indices (S-NDVI); and (3) vegetation phenological data (P-NDVI, P-EVI, R-NDVI + P-NDVI, and P-NDVI-1). Both SVM and RF classification results showed that the “R-NDVI + R-EVI” dataset achieved the highest performance, while the three pure phenological datasets produced the lowest accuracy. Correlation analysis between variable importance and rainfall time series demonstrated that the vegetation index SITS during rainfall periods showed higher importance in RF classifiers, thus revealing the potential of saving computational costs. Considering the preprocessing cost of SITS and its negative impact on the classification accuracy, we recommend overlaying the original NDVI with the original EVI time series to map the crop distribution in Kenya.

Comparative review of phenological events observed by academician P.S. Pallas in Transbaikalia in the second half of the 18th century

This article analyses the phenological data about the flowering of some plant species obtained by P.S. Pallas in Transbaikalia during his academic expedition in 1772. For seven plant species, a noticeable shift in modern flowering phases to earlier dates is shown.

Phenological and social characterization of three Lasioglossum (Dialictus) species inferred from long-term trapping collections

Detailed social and phenological data collected from nesting aggregations exist for relatively few sweat bee species because nesting aggregations are rarely found in large numbers, even when local populations are highly abundant. This limits researchers’ abilities to assess the social status of many species, which in turn, limits our ability to trace the sequence of evolutionary steps between alternative social states. To address this problem, we demonstrate the utility of rehydrated, pinned specimens from pan trap and netting collections for generating inferences about the phenology and social status of a well-studied sweat bee species, Lasioglossum (Dialictus) laevissimum. A detailed comparison of phenology and reproductive traits, between pinned specimens and those in a previous nesting study, produced similar results for bivoltine foraging activity and eusocial colony organization typical in this species. We then used pinned specimens from monitoring studies to describe, for the first time, the foraging phenology and social behaviour of two additional Dialictus species, L. hitchensi and L. ellisiae. Both L. hitchensi and L. ellisiae each exhibited two peaks in abundance during their breeding seasons, indicating two periods of foraging activity, which correspond to provisioning of spring and summer broods. Differences in body size, wear, and ovarian development of spring and summer females indicated that L. hitchensi is most likely eusocial, while L. ellisiae is either solitary or communal. This study demonstrates that analyses of specimens obtained from flower and pan trap collections can be used for assessing the phenology and social organization of temperate sweat bees in the absence of nesting data. The phenological and social lability of many sweat bee species make them ideal for studying geographic and temporal variability in sociality, and analyses of pan trap collections can make these studies possible when direct observations are impossible.

Calibration and Validation of AQUACROP and APSIM Models to Optimize Wheat Yield and Water Saving in Arid Regions

The APSIM-Wheat and AQUACROP models were calibrated for the Sakha 95 cultivar using phenological data, grain and biomass yield, and genetic parameters based on field observation. Various treatments of planting dates, irrigation, and fertilization were applied over the two successive winter growing seasons of 2019/2020 and 2020/2021. Both models simulated anthesis, maturity dates, grain yield, and aboveground biomass accurately with high performances (coefficient of determination, index of agreement greater than 0.8, and lower values of root mean square deviation) in most cases. The calibrated models were then employed to explore wheat yield and water productivity (WP) in response to irrigation and nitrogen fertilization applications. Scenario analyses indicated that water productivity and yield of wheat ranged from 1.2–2.0 kg m–3 and 6.8–8.7 t ha–1, respectively. Application of 0.8 from actual evapotranspiration and 120% from recommended nitrogen dose was the best-predicted scenario achieving the highest value of crop WP. Investigating the suitable option achieving the current wheat yield by farmers (7.4 t ha–1), models demonstrated that application of 1.4 from actual evapotranspiration with 80% of the recommended nitrogen dose was the best option to achieve this yield. At this point, predicted WP was low and recorded 1.5 kg m–3. Quantifying wheat yield in all districts of the studied area was also predicted using both models. APSIM-Wheat and AQUACROP can be used to drive the best management strategies in terms of N fertilizer and water regime for wheat under Egyptian conditions.

Ecological and biological features of silver birch (Betula pendula) in the conditions of an industrial city

The article provides data on phenology, on the content of ascorbic acid in Betula pendula Roth., growing in the conditions of the industrial city of Naberezhnye Chelny in the Republic of Tatarstan, which is a region of the Russian Federation. It is noted that natural phytocenoses are characterized by lower air temperatures in comparison with plantations of sanitary protection zones of industrial enterprises and highway plantings. According to phenological data, in urban plantings, there was observed an increase in defects, a decrease in the vital state, which is associated with severe damage to leaf blades, the formation of leaf necrosis, a decrease in the living area of leaves, which is a consequence of an intense technogenic load on woody plants. It was revealed that the content of ascorbic acid in birch leaves depends on the vegetation stage.

Phenological response to temperature variability and orography in Central Italy

Even if the sensitivity of vegetation phenology to climate change has been accepted on global and continental scales, the correlation between global warming and phenotypic variability shows a modulated answer depending on altitude, latitude, and the local seasonal thermal trend. To connect global patterns of change with local effects, we investigated the impact of the observed signal of warming found in Central Italy on two different willow species, Salix acutifolia and Salix smithiana, growing in three phenological gardens of the International Phenological Gardens’ network (IPG) located in different orographic positions. The time series of temperatures and phenological data for the period 2005–2018 were analysed first to find trends over time in the three gardens and then to correlate the recent local warming and the change in the two species phenology. The results confirmed the correlation between phenological trends and local trend of temperatures. In particular: budburst showed a trend of advancement of 1.4 days/year on average in all three gardens; flowering showed a divergent pattern between the gardens of either advancement of 1.0 days/year on average or delay of 1.1 days/year on average; while senescence showed a delay reaching even 3.3 days/year, although significant in only two gardens for both species. These trends were found to be correlated mainly with the temperatures of the months preceding the occurrence of the phase, with a shift in terms of days of the year (DOY) of the two species. Our conclusion is that the observed warming in Central Italy played a key role in controlling the phenophases occurrences of the two willow species, and that the orographic forcing leads to the different shift in DOY of phenophases (from 5 to 20 days) due to the local thermal forcing of the three phenological gardens.

Two additions to the family Malvaceae for the Flora of Telangana, India

Melhania futteyporensis Munro ex Mast and Sida sivarajanii Tambde, Sardesai & A.K. Pandey belonging to the family Malvaceae are reported here as additions to the flora of Telangana from Amrabad Tiger Reserve and Kinnerasani Wildlife Sanctuary respectively. Brief descriptions along with phenological data and colour photo plates are provided for easy identification and future reference.

Grain Yield Estimation in Rice Breeding Using Phenological Data and Vegetation Indices Derived from UAV Images

The accurate estimation of grain yield in rice breeding is crucial for breeders to screen and select qualified cultivars. In this study, a low-cost unmanned aerial vehicle (UAV) platform mounted with an RGB camera was carried out to capture high-spatial resolution images of rice canopy in rice breeding. The random forest (RF) regression techniques were used to establish yield models by using (1) only color vegetation indices (VIs), (2) only phenological data, and (3) fusion of VIs and phenological data as inputs, respectively. Then, the performances of RF models were compared with the manual observation and CERES-Rice model. The results indicated that the RF model using VIs only performed poorly for estimating yield; the optimized RF model that combined the use of phenological data and color VIs performed much better, which demonstrated that the phenological data significantly improved the model performance. Furthermore, the yield estimation accuracy of 21 rice cultivars that were continuously planted over three years in the optimal RF model had no significant difference (p > 0.05) with that of the CERES-Rice model. These findings demonstrate that the RF model, by combining phenological data and color Vis, is a potential and cost-effective way to estimate yield in rice breeding.

Machine Learning Undercounts Reproductive Organs on Herbarium Specimens but Accurately Derives Their Quantitative Phenological Status: A Case Study of Streptanthus tortuosus

Machine learning (ML) can accelerate the extraction of phenological data from herbarium specimens; however, no studies have assessed whether ML-derived phenological data can be used reliably to evaluate ecological patterns. In this study, 709 herbarium specimens representing a widespread annual herb, Streptanthus tortuosus, were scored both manually by human observers and by a mask R-CNN object detection model to (1) evaluate the concordance between ML and manually-derived phenological data and (2) determine whether ML-derived data can be used to reliably assess phenological patterns. The ML model generally underestimated the number of reproductive structures present on each specimen; however, when these counts were used to provide a quantitative estimate of the phenological stage of plants on a given sheet (i.e., the phenological index or PI), the ML and manually-derived PI’s were highly concordant. Moreover, herbarium specimen age had no effect on the estimated PI of a given sheet. Finally, including ML-derived PIs as predictor variables in phenological models produced estimates of the phenological sensitivity of this species to climate, temporal shifts in flowering time, and the rate of phenological progression that are indistinguishable from those produced by models based on data provided by human observers. This study demonstrates that phenological data extracted using machine learning can be used reliably to estimate the phenological stage of herbarium specimens and to detect phenological patterns.