A new cost‐effective approach to survey ecological communities

Surveying ecological communities often means the tedious work of collecting detailed information on each species within each sampling unit (e.g. trap, transect, quadrat). In this paper, we first argue that presence–absence and abundance data are the two extremes of a spectrum of data formats. By counting individuals of each species within a sampling unit until either a predefined (user‐defined) number of individuals is reached or all individuals of the species are counted, all intermediate cases can be generated. By independently correlating each intermediate case with the complete abundance data, we show that it is not necessary to count all individuals to recover the patterns of variation characterizing a community data table. When the same procedure is applied in combination with different distance coefficients such as the Hellinger, chord, χ², percentage difference or modified Gower, or the distance between species profiles, an even lower number of individuals per species need to be counted within a sampling unit for the patterns of variation defining a community to be recovered. By applying the same counting procedure to data collected during a pilot study, we show that the maximum number of individuals that need to be counted within a sampling unit for a species can be estimated from a pilot study containing as little as 3% of randomly selected sampling units throughout the complete survey area. An example of how to apply this new counting method is presented, using data from a boreal forest Carabidae community sampled in northwestern Alberta, Canada.     

Use of the IUL Countermat Automatic Colony Counter for Spiral Plated Total Viable Counts

An IUL Countermat automatic colony counter was used to enumerate colonies on spiral total viable count plates made with a wide variety of foods. The counter results exhibited a correlation with manual counting results similar to the reproducibility obtained with manually counted spiral plates. Use of this machine results in a large time saving compared with the conventional counting method and is recommended as a generally suitable method for counting spiral total viable count plates.     

Application of an image and environmental sensor network for automated greenhouse insect pest monitoring

This work presents an automated insect pest counting and environmental condition monitoring system using integrated camera modules and an embedded system as the sensor node in a wireless sensor network. The sensor node can be used to simultaneously acquire images of sticky paper traps and measure temperature, humidity, and light intensity levels in a greenhouse. An image processing algorithm was applied to automatically detect and count insect pests on an insect sticky trap with 93% average temporal detection accuracy compared with manual counting. The integrated monitoring system was implemented with multiple sensor nodes in a greenhouse and experiments were performed to test the system’s performance. Experimental results show that the automatic counting of the monitoring system is comparable with manual counting, and the insect pest count information can be continuously and effectively recorded. Information on insect pest concentrations were further analyzed temporally and spatially with environmental factors. Analyses of experimental data reveal that the normalized hourly increase in the insect pest count appears to be associated with the change in light intensity, temperature, and relative humidity. With the proposed system, laborious manual counting can be circumvented and timely assessment of insect pest and environmental information can be achieved. The system also offers an efficient tool for long-term insect pest behavior observations, as well as for practical applications in integrated pest management (IPM).     

DeepWILD: Wildlife Identification,Localisation and estimation on camera trap videos using Deep learning

Videos and images from camera traps are more and more used by ecologists to estimate the population of species on a territory. It is a laborious work since experts have to analyse massive data sets manually. This takes also a lot of time to filter these videos when many of them do not contain animals or are with human presence. Fortunately, deep learning algorithms for object detection can help ecologists to identify multiple relevant species on their data and to estimate their population. In this study, we propose to go even further by using object detection model to detect, classify and count species on camera traps videos. To this end, we developed a 3-step process: (i) At the first stage, after splitting videos into images, we annotate images by associating bounding boxes to each label thanks to MegaDetector algorithm; (ii) then, we extend MegaDetector based on Faster R-CNN architecture with backbone Inception-ResNet-v2 in order to not only detect the 13 relevant classes but also to classify them; (iii) finally, we design a method to count individuals based on the maximum number of bounding boxes detected. This final stage of counting is evaluated in two different contexts: first including only detection results (i.e. comparing our predictions against the right number of individuals, no matter their true class), then an evolved version including both detection and classification results (i.e. comparing our predictions against the right number in the right class). The results obtained during the evaluation of our model on the test data set are: (i) 73,92% mAP for classification, (ii) 96,88% mAP for detection with a ratio Intersection-Over-Union (IoU) of 0.5 (overlapping ratio between groundtruth bounding box and the detected one), and (iii) 89,24% mAP for detection at IoU = 0.75. Highly represented classes, like humans, have highest values of mAP around 81% whereas less represented classes in the train data set, such as dogs, have lowest values of mAP around 66%. Regarding the proposed counting method, we predicted a count either exact or $\pm$ 1 unit for 87% with detection results and for 48% with detection and classification results of our test data set. Our model is also able to detect empty videos. To the best of our knowledge, this is the first study in France about the use of object detection model on a French national park to locate, identify and estimate the population of species from camera trap videos.     

Movement patterns of selected insect groups between natural forest,open land and rubber plantation in a tropical landscape (southern Yunnan,SW China)

In many regions of tropical Asia, the expansion of rubber monoculture plantations is conducted by replacement of natural forest areas and strongly affects biodiversity and movement patterns of wild species, including insects. Against this background, we conducted a study on selected insect groups (longhorn beetles, bark beetles, wild bees and hoverflies) along transects between rainforest patches, open uncultivated land and rubber plantation habitats in a region of Xishuangbanna (southern Yunnan, China), with the objectives to identify (a) movement directions and patterns of selected insect groups based on their abundances in modified Malaise traps in the different habitats, and (b) the role of remaining natural rainforest patches and rubber plantations, respectively, for insect diversity maintenance and conservation. The highest total numbers of species and individuals of bark beetles, longhorn beetles and wild bees were recorded from the natural forest edge compared to open land and rubber plantation edge. This result clearly indicates that the natural forest plays an important role in maintenance of these three insect groups. However, the highest number of hoverfly species and individuals was recorded from the open land sites, indicating the most relevant habitat type for this group of species. Overall, the lowest species and individual numbers were recorded from the rubber plantation edge, indicating the unsuitability of this habitat type for all insect groups considered. The distribution of species and individuals in the opposite trap sides along the transect indicates that longhorn beetles, bark beetles and wild bees show not only movements from the forest to the surrounding habitats, but also return back after encountering the unsuitable rubber plantation habitat. Bark beetle composition showed the relatively highest similarity between all trap sites and opposite trap sides among the insect groups considered, indicating a higher movement activity than the other groups. The four insect groups considered in this study show different movement modes between the forest, open land and rubber plantation, which are not the same for all taxa. Except for hoverflies, the natural forest was found to be the most important habitat for the maintenance of species diversity in the different land use types of the study area.     

转基因抗虫水稻对水稻害虫群落的影响

本文应用吸虫器法与剥查法,研究了转Cry1Ac +SCK基因水稻及其杂交后代对水稻害虫群落的影响。结果显示：转Cry1Ac+SCK基因水稻MSA与MSB对稻纵卷叶螟具有极强的抗性,转Cry1Ac+SCK基因稻的杂交后代21S/MSB与Ⅱ32A/MSB都获得了对稻纵卷叶螟的高抗性。虽然MSA、MSB、21S/MSB与Ⅱ32A/MSB对水稻害虫的物种丰富度、个体数量、多样性与均匀度时间动态都有一定程度的影响,但生长全期的调查结果显示,与对应的对照比较,它们对水稻害虫群落的这些参数均无显著影响,且转基因水稻田与各自对照田在害虫与主要害虫物种组成上具有较高的相似性。靶标害虫种群数量的降低,并未对水稻害虫群落的组成与结构产生明显的影响,也没有引起优势种成分的变化。     

The Amazonas‐trap: a new method for sampling plant‐inhabiting arthropod communities in tropical forest understory

Methods to quantify plant‐insect interactions in tropical forests may miss many important arthropods and can be time consuming and uneven in capture efficiency. We describe the Amazonas‐trap, a new method that rapidly envelops the target plant for sampling arthropods. We evaluated the efficiency of the Amazonas‐trap by comparing it with two commonly used sampling methods to collect arthropods from plants: the beating tray and manual collection. Samples were collected in 10 permanent plots, in the Ducke forest reserve, Manaus (Amazonas, Brazil). In each plot we sampled 18 plant individuals of Protium sp. (Burseraceae): six by a beating tray, six by manual collection, and six using the Amazonas‐trap. All insects were identified to the family level and those belonging to the order Hymenoptera were identified to the species and morphospecies level. The new method sampled more insect families and more Hymenoptera species than tree beating and manual collection. Of the 75 total families collected, 20 were sampled exclusively by the Amazonas‐trap, seven were only collected with a beating tray, and seven were sampled exclusively with manual collecting. A similar pattern was found for abundance: Amazonas‐trap sampled more individuals, followed by the beating tray and manual collection. Small and winged arthropods were more abundant in Amazonas‐trap, explaining the highest richness of Hymenoptera and insect families sampled with this method. The new method sampled more spiders, wood‐fungi feeders, sap suckers, omnivorous, parasitoids, and insect predators than the other methods, but was equally effective in sampling leaf‐feeders and ants. Amazonas‐trap was more time consuming in the field, but for all diversity parameters evaluated, the new method showed better performance for collecting invertebrates on plants.     

Comparison of four trap types for ambrosia beetles (Coleoptera, Scolytidae) in Brazilian Eucalyptus stands

Eucalyptus spp. plantations represent >60% of the reforested area in Brazil. Although ambrosia beetle attacks on live trees were at first nonexistent, they have begun to appear with greater frequency. Monitoring for pest insects is a key factor in integrated pest management, and baited traps are one of the most widely used methods for insect population detection and survey. We compared the efficiency of the most widely used trap in Brazil to survey for ambrosia beetles and other Scolytidae, the ESALQ-84 type, with other traditionally employed traps: the multiple funnel (Lindgren trap); drainpipe; and slot (Theyson) traps, in a Eucalyptus grandis Hill ex Maiden stand in Brazil. The ESALQ-84 trap was the most efficient in trapping Hypothenemus eruditus Westwood and Hypothenemus obscurus (F.); the multiple funnel trap caught significantly more Cryptocarenus diadematus Eggers; whereas the slot trap caught more Premnobius cavipennis Eichhoff and Xyleborus affinis Eichhoff than the other traps. The drainpipe trap was the least effective trap overall. When corrected for number of beetles caught per trap surface area, catches were significantly higher on the ESALQ-84 trap for the majority of the species analyzed, probably because of a smaller trap surface area. The slot trap was recommended for it caught overall more beetles of the three most economically important scolytid species in eucalypt plantations in Brazil, P. cavipennis, X. affinis, and X. ferrugineus.     

QUANTITATIVE METHOD FOR DETERMINING A REPRESENTATIVE ALGAL SAMPLE COUNT1

A method for determining a representative count of a sample dependent on number of species is presented for application to various algal communities. Constant species curves are calculated as efficiency = (number of individuals–number of species)/number of individuals and diagrammed on a plot of efficiency versus number of individuals counted. Efficiency is defined as the probability that a new species encountered is minimal. That is, as the ratio of number of species to number of individuals approaches 1, more individuals will need to be counted in order to achieve a representative count. Data and calculations of efficiency from two algal communities are presented for illustration.     

Monitoring stored-product pests in food processing plants with pheromone trapping,contour mapping,and mark-recapture

Distribution and movement patterns of several species of stored-product pests in a food processing plant were investigated. The objectives of this study were to determine the temporal and spatial variation in abundance of stored-product pests using pheromone traps; assess the effectiveness of trap type, location, and number on monitoring insect populations; and to evaluate the nature of pheromone trap capture hot spots by measuring patterns of insect movement. We determined that the distributions of Trogoderma variabile Ballion, Lasioderina serricorne (F.), Tribolium castaneum (Herbst), and Plodia interpunctella (Hübner) within the facility were typically clumped and that foci of high trap captures, based on visual observation of contour maps, varied among species and over time. Trap type and location influenced the number of T. variabile captured: traps on the floor and along walls captured more individuals than hanging traps and traps next to support pillars. T. variabile was the predominant insect pest at this facility and from mark-recapture studies, we found that individual beetles moved across multiple floors in the facility and from 7 to 216 m though the warehouse.     

基于林分昆虫多样性调查的传统灯诱法改进效果分析

昆虫在生物多样性调查中具有不可替代的地位,灯诱法是效率最高的昆虫采集方法之一,然而传统的灯诱方法存在诱集范围无法界定、采样流程标准性不足等问题。为此,使用自制的聚光灯诱箱（Spotlight Box Trap,SBT）作为改进灯诱方法,并使用灯诱帐（Moth Collecting Tent,MCT）作为传统灯诱法进行对照;以北京杨（Populus×beijingensis）与油松（Pinus tabuliformis）人工林为样地,对两种灯诱装置采集的蛾类数量及物种组成、物种多样性指数与形态参数等多个指标进行评价。共采集蛾类24科164种6728号,其中北京杨林分SBT采集19科93种1164号,MCT采集22科125种2573号;油松林分SBT采集21科87种1044号,MCT采集22科117种1947号。分析显示：（1） MCT采集的个体数量与物种数量均高于SBT,总体上SBT在不同样地间的数据误差水平更低。（2）两种灯诱装置采集的蛾类优势种组成显著不同（P<0.05）;SBT共有种的数量比例始终高于MCT,即后者的特有种占比例更高。（3）总体上SBT物种多样性指数的数据误差水平更低。（4） MCT采集的蛾类物种体型更大（体长约为SBT的1.20倍）、飞行能力更强（相关参数约为SBT的1.20倍至1.36倍）。结果表明,在林分昆虫多样性的调查中,传统灯诱法的数据稳定性较差,对林分的代表性不足。聚光灯诱箱因其对光域的可控性和采集昆虫的规范性,具备更高的稳定性,在森林昆虫多样性调查与监测中具有更好的适用性与应用前景。     

How varying pest and trap densities affect Tribolium castaneum capture in pheromone traps

The red flour beetle, Tribolium castaneum (Herbst) (Coleoptera: Tenebrionidae), is an important insect pest in food processing facilities. Pheromone trapping is frequently used to monitor red flour beetle populations in structures; however, the optimal trap density and the relationship between trap captures and beetle density is not known. Two experiments were performed concurrently in environmentally controlled 30‐m² walk‐in chambers to determine the relationship between aggregation pheromone trap captures of red flour beetles and beetle and trap number. In one experiment, beetle density was kept constant at 200 individuals per chamber while trap number was varied from 1 to 8, and in the other experiment trap number remained constant at one per chamber while beetle density varied from 20 to 800 individuals. Results indicated that approximately one out of 23 red flour beetles were captured in a trap. Number of beetles captured in traps increased significantly as beetle density increased; however, the proportion of beetles captured remained consistent across beetle densities with a mean of 4.7 ± 0.6% of individuals captured. Trap captures varied significantly with trap placement within experimental chambers, indicating that subtle differences in the trapping environment can influence trap captures. Data suggested that trap densities of 0.07–0.10 m^?2 (2–3 traps per chamber) would maximize trap capture, whereas a trap density of 0.13 m^?2 (four traps per chamber) would maximize the predictive ability of a trapping equation estimating beetle density from trap captures. Results provide information needed to more thoroughly explore how environmental factors might influence red flour beetle trap capture in the absence of changes in beetle density. Further understanding of these relationships will allow for more accurate assessments of absolute beetle density from pheromone trap capture data.     

Species‐level image classification with convolutional neural network enables insect identification from habitus images

1. Changes in insect biomass, abundance, and diversity are challenging to track at sufficient spatial, temporal, and taxonomic resolution. Camera traps can capture habitus images of ground‐dwelling insects. However, currently sampling involves manually detecting and identifying specimens. Here, we test whether a convolutional neural network (CNN) can classify habitus images of ground beetles to species level, and estimate how correct classification relates to body size, number of species inside genera, and species identity.
2. We created an image database of 65,841 museum specimens comprising 361 carabid beetle species from the British Isles and fine‐tuned the parameters of a pretrained CNN from a training dataset. By summing up class confidence values within genus, tribe, and subfamily and setting a confidence threshold, we trade‐off between classification accuracy, precision, and recall and taxonomic resolution.
3. The CNN classified 51.9% of 19,164 test images correctly to species level and 74.9% to genus level. Average classification recall on species level was 50.7%. Applying a threshold of 0.5 increased the average classification recall to 74.6% at the expense of taxonomic resolution. Higher top value from the output layer and larger sized species were more often classified correctly, as were images of species in genera with few species.
4. Fine‐tuning enabled us to classify images with a high mean recall for the whole test dataset to species or higher taxonomic levels, however, with high variability. This indicates that some species are more difficult to identify because of properties such as their body size or the number of related species.
5. Together, species‐level image classification of arthropods from museum collections and ecological monitoring can substantially increase the amount of occurrence data that can feasibly be collected. These tools thus provide new opportunities in understanding and predicting ecological responses to environmental change.

     

A novel concentration and viability detection method for <Emphasis Type="Italic">Brettanomyces</Emphasis> using the Cellometer image cytometry

Brettanomyces spp. can present unique cell morphologies comprised of excessive pseudohyphae and budding, leading to difficulties in enumerating cells. The current cell counting methods include manual counting of methylene blue-stained yeasts or measuring optical densities using a spectrophotometer. However, manual counting can be time-consuming and has high operator-dependent variations due to subjectivity. Optical density measurement can also introduce uncertainties where instead of individual cells counted, an average of a cell population is measured. In contrast, by utilizing the fluorescence capability of an image cytometer to detect acridine orange and propidium iodide viability dyes, individual cell nuclei can be counted directly in the pseudohyphae chains, which can improve the accuracy and efficiency of cell counting, as well as eliminating the subjectivity from manual counting. In this work, two experiments were performed to demonstrate the capability of Cellometer image cytometer to monitor Brettanomyces concentrations, viabilities, and budding/pseudohyphae percentages. First, a yeast propagation experiment was conducted to optimize software counting parameters for monitoring the growth of Brettanomyces clausenii, Brettanomyces bruxellensis, and Brettanomyces lambicus, which showed increasing cell concentrations, and varying pseudohyphae percentages. The pseudohyphae formed during propagation were counted either as multiple nuclei or a single multi-nuclei organism, where the results of counting the yeast as a single multi-nuclei organism were directly compared to manual counting. Second, a yeast fermentation experiment was conducted to demonstrate that the proposed image cytometric analysis method can monitor the growth pattern of B. lambicus and B. clausenii during beer fermentation. The results from both experiments displayed different growth patterns, viability, and budding/pseudohyphae percentages for each Brettanomyces species. The proposed Cellometer image cytometry method can improve efficiency and eliminate operator-dependent variations of cell counting compared with the traditional methods, which can potentially improve the quality of beverage products employing Brettanomyces yeasts.     

Automatic cell counting with ImageJ

Cell counting is an important routine procedure. However, to date there is no comprehensive, easy to use, and inexpensive solution for routine cell counting, and this procedure usually needs to be performed manually. Here, we report a complete solution for automatic cell counting in which a conventional light microscope is equipped with a web camera to obtain images of a suspension of mammalian cells in a hemocytometer assembly. Based on the ImageJ toolbox, we devised two algorithms to automatically count these cells. This approach is approximately 10 times faster and yields more reliable and consistent results compared with manual counting.     

A model-based high throughput method for fecundity estimation in fruit fly studies

ABSTRACT

The ability to quantify fecundity is critically important to a wide range of experimental applications, particularly in widely-used model organisms such as Drosophila melanogaster. However, the standard method of manually counting eggs is time consuming and limits the feasibility of large-scale experiments. We develop a predictive model to automate the counting of eggs from images of eggs removed from the media surface and washed onto dark filter paper. Our method uses the simple relationship between the white area in an image and the number of eggs present to create a predictive model that performs well even at high egg densities where clumping can complicate the individual identification of eggs. A cross-validation approach demonstrates our method performs well, with a correlation between predicted and manually counted values of 0.88. We show how this method can be applied to a large data set where egg densities vary widely.     

Quick and efficient measurement of transient GUS expression using image analysis

An image-analysis procedure was developed for measurement of transient β-glucuronidase (GUS) gene expression events in peanut. Gene transfer via microprojectile bombardment has been effective in peanut. Experiments were performed to improve the efficiency of DNA delivery to embryogenic calli for transformation. Transient GUS expression was assayed 24 hours after bombardment by staining calli and manually counting blue-stained spots under a microscope. These calli were then placed under a dissecting microscope with a video camera that recorded each image onto a computer. The images were then subjected to analysis with software that counted the number of blue-stained GUS-positive spots as well as the area and percent of the total area of each spot. The manual counts and data from the image analysis were compared through correlation statistics and by comparing analyses of variance for fixed parameters of the experiments. Correlation coefficients were high between manual spot counts and image-analysis counts (r=0.77) and between manual counts and percent area (GUS (r=0.91) Results indicate that image analysis provides a quick and efficient method for analysis of transient expression. Much more reliable information is obtained with less effort and less fatigue with the image-analysis procedure. Scientists can use this method to monitor transient expression events in plant transformation experiments.     

Suitability of commercially available insect traps and pheromones for monitoring dusky sap beetles (Coleoptera: Nitidulidae) and related insects in Bt sweet corn

Two trap types and pheromone sources for the dusky sap beetle, Carpophilus lugubris Murray (Coleoptera: Nitidulidae), were compared in Bacillus thuringiensis (Bt) and non-Bt sweet corn fields over a 3-yr period. Overall, commercial traps and pheromones were equally effective as experimental traps and pheromones used previously for capturing C. lugubris and other sap beetle species. The commercial trap often caught significantly more Glischrochilus quadrisignatus Say than the experimental trap that had been used in previous studies. Bt corn significantly reduced caterpillar damage to ears compared with the non-Bt isoline and did not adversely affect levels of Orius sp., the most common insect predator. Sap beetle damage was the most common insect damage to Bt sweet corn ears. Sap beetles were detected by traps at population levels below that which are likely to cause economic concern, indicating commercially available traps and pheromone lures for monitoring sap beetles should be suitable for detecting them under commercial growing conditions.     

A method for automatic real-time detection and counting of fruit fly pests in orchards by trap bottles via convolutional neural network with attention mechanism added

Fruit fly pests seriously affect the quality and safety of various melons, fruits, and vegetable crops. Many farmers lack sufficient knowledge of the level of pest occurrence, which leads to the over-use of pesticides, resulting in environmental pollution and quality degradation of agricultural products. The combination of artificial intelligence and agricultural technology can continuously and dynamically monitor pests in orchards, help scientific researchers and fruit farmers master pest data in time, reduce the use of artificial and pesticides, and achieve scientific early warning and prevention of pests. In this paper, the sexual attractant is placed in the pest trap bottle, the optical flow method, U-Net semantic segmentation, and the YOLOv5 algorithm with SE, CBAM, CA, and ECA attention mechanisms are used to detect and count live Bactrocera cucurbitae on the bottle surface, and then we use Hough circle detection method to detect the entrance position of the trap bottle, and finally count the number of B. cucurbitae entering the trap bottle in combination with the position information of B. cucurbitae and the entrance of the trap bottle. The experimental results show that: the accuracy of counting B. cucurbitae on the surface of the trap bottle can reach 93.5%, and the accuracy of counting B. cucurbitae entering the trap bottle can reach 94.3%, which can dynamically monitor the number of pests in the orchard in real time.     

Feeding Fauna and Foraging Habits of Tiger Beetles Found in Agro‐ecosystems in Western Ghats,India1

Cicindela (Calochroa) whithilli (Hope) and Cicindela (Calochroa) flavomaculata Hope (Cicindelidae: Coleoptera) are seasonally dominant predatory insects in the cultivated, and irrigated rice paddy fields of the South Indian peninsula. While studying the feeding ecology of these tiger beetles in rice paddy agro‐ecosystems in Sringeri area of the central Western Ghats, we examined their potential as biocontrols on the major rice paddy pest populations available in this region. Earthworms and tadpoles were significant prey organisms during the early cultivation period, while other traditional prey organisms like ants and spiders (mainly wolf and jumping spiders) were more common as prey organisms toward the mid‐season. The results showed that the feeding fauna of tiger beetles can extend to prey items beyond arthropods, to include vertebrates. Cicindela (Calochroa) duponti Dejean is another common species that occurs abundantly in the Areca orchards and on the bunds that separate rice paddy fields from adjacent uncultivated lands. We performed laboratory tests to determine the feeding guild of these beetles, and to investigate their possible role as a predator of the rice paddy pests. Although preliminary in nature, the results suggest that both C. whithilli and C. flavomaculata are ineffective as biocontrol agents of rice paddy agro‐ecosystems. They appear to have only a marginal impact on the larvae of Leptocorisa acuta, a major insect pest of paddy rice, but further experimental and observational studies are needed to firmly establish the significance of this observation.