Drivers of reef fish assemblages in an upwelling region from the Eastern Tropical Pacific Ocean

Reef fish assemblages are exposed to a wide range of anthropogenic threats as well as chronic natural disturbances. In upwelling regions, for example, there is a seasonal influx of cool nutrient-rich waters that may shape the structure and composition of reef fish assemblages. Given that climate change may disrupt the natural oceanographic processes by altering the frequency and strength of natural disturbances, understanding how fish assemblages respond to upwelling events is essential to effectively manage reef ecosystems under changing ocean conditions. This study used the baited remote underwater video stations (BRUVS) and the traditional underwater visual census (UVC) to investigate the spatiotemporal patterns of reef fish assemblages in an upwelling region in the North Pacific of Costa Rica. A total of 183 reef fish species from 60 families were recorded, of which 166 species were detected using BRUVS and 122 using UVC. Only 66% of all species were detected using both methods. This study showed that the upwelling had an important role in shaping reef fish assemblages in this region, but there was also a significant interaction between upwelling and location. In addition, other drivers such as habitat complexity and habitat composition had an effect on reef fish abundances and species. To authors’ knowledge, this is the first study in the Eastern Tropical Pacific that combines BRUVS and UVC to monitor reef fish assemblages in an upwelling region, which provides more detailed information to assess the state of reef ecosystems in response to multiple threats and changing ocean conditions.     

CO-WOA: Novel Optimization Approach for Deep Learning Classification of Fish Image

The most significant groupings of cold-blooded creatures are the fish family. It is crucial to recognize and categorize the most significant species of fish since various species of seafood diseases and decay exhibit different symptoms. Systems based on enhanced deep learning can replace the area's currently cumbersome and sluggish traditional approaches. Although it seems straightforward, classifying fish images is a complex procedure. In addition, the scientific study of population distribution and geographic patterns is important for advancing the field's present advancements. The goal of the proposed work is to identify the best performing strategy using cutting-edge computer vision, the Chaotic Oppositional Based Whale Optimization Algorithm (CO-WOA), and data mining techniques. Performance comparisons with leading models, such as Convolutional Neural Networks (CNN) and VGG-19, are made to confirm the applicability of the suggested method. The suggested feature extraction approach with Proposed Deep Learning Model was used in the research, yielding accuracy rates of 100 %. The performance was also compared to cutting-edge image processing models with an accuracy of 98.48 %, 98.58 %, 99.04 %, 98.44 %, 99.18 % and 99.63 % such as Convolutional Neural Networks, ResNet150V2, DenseNet, Visual Geometry Group-19, Inception V3, Xception. Using an empirical method leveraging artificial neural networks, the Proposed Deep Learning model was shown to be the best model.     

Macrophytes in the Upper Paraná River floodplain: checklist and comparison with other large South American wetlands

Neotropical aquatic ecosystems have a rich aquatic flora. In this report, we have listed the aquatic flora of various habitats of the upper Paraná River floodplain by compiling data from literature and records of our own continuous collections conducted during the period 2007-2009. Our main purposes were to assess the macrophyte richness in the Paraná floodplain, to compare it with other South American wetlands and to assess whether the number of species recorded in South American inventories has already reached an asymptote. We recorded a total of 153 species of macrophytes in the Upper Paraná River floodplain, belonging to 100 genera and 47 families. In our comparative analysis, a clear floristic split from other South American wetlands was shown, except for the Pantanal, which is the closest wetland to the Paraná floodplain and, therefore, could be considered a floristic extension of the Pantanal. The species accumulation curve provides evidence that sampling efforts should be reinforced in order to compile a macrophyte flora census for South America. The high dissimilarity among South American wetlands, together with the lack of an asymptote in our species accumulation curve, indicates that the sampling effort needs to be increased to account for the actual species richness of macrophytes in this region.     

Classification of desert grassland species based on a local-global feature enhancement network and UAV hyperspectral remote sensing

Grassland ecosystems are an important part of terrestrial ecosystems and are important for building ecological barriers, promoting the pastoral economy, and maintaining social stability. In recent decades, grasslands in northern China have undergone extensive degradation due to the combined effects of global climate change and the anthropogenic overuse of grasslands. An understanding of the spatial distribution of grassland degradation species is helpful for evaluating the process of grassland degradation and formulating appropriate protective measures. This is important for grassland degradation monitoring. To address the limitations of traditional ground surveys and realize intelligent remote sensing grassland degradation monitoring tasks, we use unmanned aerial vehicle (UAV) hyperspectral remote sensing technology to collect data on vegetation species in desert grasslands. In this paper, we propose a local-global feature enhancement network (LGFEN) for the classification of desert grassland species. The method uses the local feature enhancement (LFE) module and global feature enhancement (GFE) module to extract local and spatial information from hyperspectral images (HSIs), respectively. In addition, the introduction of the convolutional block attention module (CBAM) refines the features of HSIs, improving the stability of the classification performance. The results show that the proposed method has superior classification performance compared with existing HSI classification methods. With only 10 training samples per class, the overall accuracy, average accuracy, and kappa coefficient of the proposed method were 98.61%, 97.61%, and 0.9815, respectively. The proposed method provides a new approach for high-precision and high-efficiency dynamic monitoring of grassland ecosystems.     

Computer vision technique for freshness estimation from segmented eye of fish image

Preserving the quality of fish is a challenging task. Several different cooling methods and materials are used during their storage, transportation purpose. These are responsible factors that decide the freshness of a post harvested fish. In this proposed algorithm, a computer vision-based technique is developed to predict the freshness level of fish from its image. Eyes of the fish are considered as the region of interest, as a good correlation has been observed between the colour of the eye and different duration of storage day. It is segmented from the image of a fish sample and then a strategic framework is used for extraction of the discriminatory features. These extracted features show a degradation pattern which acts as an indicative parameter to determine the level of freshness of sample of fish. The proposed method provides a recognition accuracy of 96.67%. The experimental results indicate that this is an efficient and non-destructive methodology for detecting the fish freshness. The high accuracy of freshness detection and low computation time makes this non-destructive methodology efficient for real-world usage in the fish industry and market.     

Genetic variation and population structure of the endangered Hyacinth Macaw (<Emphasis Type="Italic">Anodorhynchus hyacinthinus</Emphasis>): implications for conservation

The Hyacinth Macaw (Anodorhynchus hyacinthinus) is one of 14 endangered species in the family Psittacidae occurring in Brazil, with an estimated total population of 6,500 specimens. We used nuclear molecular markers (single locus minisatellites and microsatellites) and 472 bp of the mitochondrial DNA control region to characterize levels of genetic variability in this species and to assess the degree of gene flow among three nesting sites in Brazil (Pantanal do Abobral, Pantanal de Miranda and Piauí). The origin of five apprehended specimens was also investigated. The results suggest that, in comparison to other species of parrots, Hyacinth Macaws possess relatively lower genetic variation and that individuals from two different localities within the Pantanal (Abobral and Miranda) belong to a unique interbreeding population and are genetically distinct at nuclear level from birds from the state of Piauí. The analyses of the five apprehended birds suggest that the Pantanal is not the source of birds for illegal trade, but their precise origin could not be assigned. The low genetic variability detected in the Hyacinth Macaw does not seem to pose a threat to the survival of this species. Nevertheless, habitat destruction and nest poaching are the most important factors negatively affecting their populations in the wild. The observed genetic structure emphasizes the need of protection of Hyacinth Macaws from different regions in order to maintain the genetic diversity of this species.     

Identification of blood species based on surface-enhanced Raman scattering spectroscopy and convolutional neural network

The identification of blood species is of great significance in many aspects such as forensic science, wildlife protection, and customs security and quarantine. Conventional Raman spectroscopy combined with chemometrics is an established method for identification of blood species. However, the Raman spectrum of trace amount of blood could hardly be obtained due to the very small cross-section of Raman scattering. In order to overcome this limitation, surface-enhanced Raman scattering (SERS) was adopted to analyze trace amount of blood. The 785 nm laser was selected as the optimal laser to acquire the SERS spectra, and the blood SERS spectra of 19 species were measured. The convolutional neural network (CNN) was used to distinguish the blood of 19 species including human. The recognition accuracy of the blood species was obtained with 98.79%. Our study provides an effective and reliable method for identification and classification of trace amount of blood.     

Diversity and effect of historical inundation on bee and wasp (Hymenoptera: Apoidea,Vespoidea) communities in the Brazilian Pantanal

Hymenoptera are the third largest insect order and is one of the most important ecological agents in terrestrial ecosystems, while terrestrial flooded areas are considered priority for conservation due to their uniqueness. We present the first inventory for bees (Apoidea) and wasps (Vespoidea) and describe their alpha, beta, and gamma diversity s in the Brazilian Pantanal to evaluate the effect of the flood intensity on the community structure. We tested the hypothesis that areas with different profiles would present less similar composition, for this, sampling was carried out in the Brazilian Pantanal between November 2015 and March 2016 using passive traps and active search with entomological net in 19 areas distributed in five flood categories. Ecological metrics were used to describe the community, as well as multivariate analyzes for community interpretation. We collected 3342 individuals belonging to 377 species and eight families and expected 532 (±?23) species of Apoidea and Vespoidea. Different areas presented low similarity and the compositions of the communities tended to be different, mainly between the extremes of the flood profile for different families. The biological characteristics of representatives from the highest taxonomic levels (Subfamily, Tribe) are important for occupation of the areas in the Pantanal. Describing the fauna and understanding how it is influenced by flooding history provides tools for conservation strategies in the Pantanal and further effort is needed to preserve distinct areas within the biome.     

Growth parameters estimates for a small fish of the Pantanal, Brazil: Moenkhausia dichroura (Characiformes; Characidae).

Growth parameters were estimated for Moenkhausia dichroura (Kner, 1858) (Characiformes, Characidae), a small-sized and very abundant fish of the Pantanal lentic habitats commonly known as "pequira ". A method based on the length frequencies distribution and the ELEFAN I routine from the FISAT program were used. The fish were collected in the Baia da On?a, an oxbowlake of the sub-region Pantanal of Aquidauana, Mato Grosso do Sul, Brazil, from June to December 1988. The standard length of the captured fishes ranged from 29 to 76 mm with an average of 53 mm. The estimated growth parameters were L(infinity) = 81 mm (standard length), k = 0.85 year(-1), C = 0.89, WP = 0.6 (Rn = 0.285). The WP indicated that growth reduction occurred in July, when the lowest temperature of the year was registered. The growth curve showed that captured individuals belonged to three cohorts. The obtained results seem robust and quite compatible with the biology of the fish and its adjustment to the environment. M. dichroura, in spite of not being a direct fishing interest, is an important species in terms of its ecological aspects, due to its abundance and high growth rate, and as a great food source for aquatic organisms and specially for larger fish of economic value. Considering the information gap about small fish, the parameters estimated for pequira constitute a comparison base for other growth studies of small-sized fish species of tropical environments.     

全景绿视率自动识别和计算研究

绿视率是用于绿色空间感知的直观评价标准,传统研究的绿视率多基于平面影像进行计算,不能完全反映三维空间中人对绿量的主观感受。基于全景影像,提出全景绿视率的概念,通过全景相机获取球面全景照片,将等距圆柱投影转换为等积圆柱投影,利用基于语义分割的卷积神经网络模型,自动识别植被区域面积以实现全景绿视率自动化识别和计量。通过比较5项卷积神经网络模型对绿视率的识别效果,显示出Dilated ResNet-105神经网络模型具有最高的识别准确度。以武汉市武昌区紫阳公园为例,对各级园路和广场的全景绿视率进行计算和分析。将卷积神经网络的识别结果同人工判别结果进行对比研究,结果显示：使用Dilated ResNet-105卷积神经网络对绿植范围识别的平均交并比（mIoU）为62.53%,与人工识别的平均差异为9.17%。全景绿视率自动识别和计算可以为相关研究提供新的思路,实现客观准确、快速便捷的绿视率测量评估。     

Landscape‐scale determinants of non‐native fish communities

Aim Predicting and preventing invasions depends on knowledge of the factors that make ecosystems susceptible to invasion. Current studies generally rely on non‐native species richness (NNSR) as the sole measure of ecosystem invasibility; however, species identity is a critical consideration, given that different ecosystems may have environmental characteristics suitable to different species. Our aim was to examine whether non‐native freshwater fish community composition was related to ecosystem characteristics at the landscape scale. Location United States. Methods We described spatial patterns in non‐native freshwater fish communities among watersheds in the Mid‐Atlantic region of the United States based on records of establishment in the U.S. Geological Survey’s Nonindigenous Aquatic Species Database. We described general relationships between non‐native species and ecosystem characteristics using canonical correspondence analysis. We clustered watersheds by non‐native fish community and described differences among clusters using indicator species analysis. We then assessed whether non‐native communities could be predicted from ecosystem characteristics using random forest analysis and predicted non‐native communities for uninvaded watersheds. We estimated which ecosystem characteristics were most important for predicting non‐native communities using conditional inference trees. Results We identified four non‐native fish communities, each with distinct indicator species. Non‐native communities were predicted based on ecosystem characteristics with an accuracy of 80.6%, with temperature as the most important variable. Relatively uninvaded watersheds were predicted to be invasible by the most diverse non‐native community. Main conclusions Non‐native species identity is an important consideration when assessing ecosystem invasibility. NNSR alone is an insufficient measure of invasibility because ecosystems with equal NNSR may not be equally invasible by the same species. Our findings can help improve predictions of future invasions and focus management and policy decisions on particular species in highly invasible ecosystems.     

Automatic classification of Furnariidae species from the Paranaense Littoral region using speech-related features and machine learning

Over the last years, researchers have addressed the automatic classification of calling bird species. This is important for achieving more exhaustive environmental monitoring and for managing natural resources. Vocalisations help to identify new species, their natural history and macro-systematic relations, while computer systems allow the bird recognition process to be sped up and improved. In this study, an approach that uses state-of-the-art features designed for speech and speaker state recognition is presented. A method for voice activity detection was employed previous to feature extraction. Our analysis includes several classification techniques (multilayer perceptrons, support vector machines and random forest) and compares their performance using different configurations to define the best classification method. The experimental results were validated in a cross-validation scheme, using 25 species of the family Furnariidae that inhabit the Paranaense Littoral region of Argentina (South America). The results show that a high classification rate, close to 90%, is obtained for this family in this Furnariidae group using the proposed features and classifiers.     

A robust deep attention dense convolutional neural network for plant leaf disease identification and classification from smart phone captured real world images

Plant diseases cause significant food loss and hence economic loss around the globe. Therefore, automatic plant disease identification is a primary task to take proper medications for controlling the spread of the diseases. Large variety of plants species and their dissimilar phytopathological symptoms call for the implementation of supervised machine learning techniques for efficient and reliable disease identification and classification. With the development of deep learning strategies, convolutional neural network (CNN) has paved its way for classification of multiple plant diseases by extracting rich features. However, several characteristics of the input images especially captured in real world environment, viz. complex or indistinguishable background, presence of multiple leaves with the diseased leaf, small lesion area, solemnly affect the robustness and accuracy of the CNN modules. Available strategies usually applied standard CNN architectures on the images captured in the laboratory environment and very few have considered practical in-field leaf images for their studies. However, those studies are limited with very limited number of plant species. Therefore, there is need of a robust CNN module which can successfully recognize and classify the dissimilar leaf health conditions of non-identical plants from the in-field RGB images. To achieve the above goal, an attention dense learning (ADL) mechanism is proposed in this article by merging mixed sigmoid attention learning with the basic dense learning process of deep CNN. The basic dense learning process derives new features at higher layer considering all lower layer features and that provides fast and efficient training process. Further, the attention learning process amplifies the learning ability of the dense block by discriminating the meaningful lesion portions of the images from the background areas. Other than adding an extra layer for attention learning, in the proposed ADL block the output features from higher layer dense learning are used as an attention mask to the lower layers. For an effective and fast classification process, five ADL blocks are stacked to build a new CNN architecture named DADCNN-5 for obtaining classification robustness and higher testing accuracy. Initially, the proposed DADCNN-5 module is applied on publicly available extended PlantVillage dataset to classify 38 different health conditions of 14 plant species from 54,305 images. Classification accuracy of 99.93% proves that the proposed CNN module can be used for successful leaf disease identification. Further, the efficacy of the DADCNN-5 model is checked after performing stringent experiments on a new real world plant leaf database, created by the authors. The new leaf database contains 10,851 real-world RGB leaf images of 17 plant species for classifying their 44 distinguished health conditions. Experimental outcomes reveal that the proposed DADCNN-5 outperforms the existing machine learning and standard CNN architectures, and achieved 97.33% accuracy. The obtained sensitivity, specificity and false positive rate values are 96.57%, 99.94% and 0.063% respectively. The module takes approximately 3235 min for training process and achieves 99.86% of training accuracy. Visualization of Class activation mapping (CAM) depicts that DADCNN-5 is able to learn distinguishable features from semantically important regions (i.e. lesion regions) on the leaves. Further, the robustness of the DADCNN-5 is established after experimenting with augmented and noise contaminated images of the practical database.     

Unlocking biodiversity and conservation studies in high‐diversity environments using environmental DNA (eDNA): A test with Guianese freshwater fishes

Determining the species compositions of local assemblages is a prerequisite to understanding how anthropogenic disturbances affect biodiversity. However, biodiversity measurements often remain incomplete due to the limited efficiency of sampling methods. This is particularly true in freshwater tropical environments that host rich fish assemblages, for which assessments are uncertain and often rely on destructive methods. Developing an efficient and nondestructive method to assess biodiversity in tropical freshwaters is highly important. In this study, we tested the efficiency of environmental DNA (eDNA) metabarcoding to assess the fish diversity of 39 Guianese sites. We compared the diversity and composition of assemblages obtained using traditional and metabarcoding methods. More than 7,000 individual fish belonging to 203 Guianese fish species were collected by traditional sampling methods, and ~17 million reads were produced by metabarcoding, among which ~8 million reads were assigned to 148 fish taxonomic units, including 132 fish species. The two methods detected a similar number of species at each site, but the species identities partially matched. The assemblage compositions from the different drainage basins were better discriminated using metabarcoding, revealing that while traditional methods provide a more complete but spatially limited inventory of fish assemblages, metabarcoding provides a more partial but spatially extensive inventory. eDNA metabarcoding can therefore be used for rapid and large‐scale biodiversity assessments, while at a local scale, the two approaches are complementary and enable an understanding of realistic fish biodiversity.     

A lightweight CNN-based model for early warning in sow oestrus sound monitoring

The reproductive performance of sows is an important indicator for evaluating the economic efficiency and production level of pigs. In this paper, we design and propose a lightweight sow oestrus detection method based on acoustic data and deep convolutional neural network (CNN) algorithms by collecting and analysing short-frequency and long-frequency sow oestrus sounds. We use visual log-mel spectrograms, which can reflect three-dimensional information, as inputs to the network model to improve the overall recognition accuracy. The improved lightweight MobileNetV3_esnet model is used to identify oestrus and nonoestrus sounds and is compared with existing algorithms. The model outperforms the other algorithms, with 97.12% precision, 97.34% recall, 97.59% F1-score, and 97.52% accuracy; the model size is 5.94 MB. Compared with traditional oestrus monitoring methods, the proposed method can more accurately boost the vocal characteristics exhibited by sows in latent oestrus, thus providing an efficient and accurate approach for use in practical applications of oestrus monitoring and early warning systems on pig farms.     

Sustainable harvest of two large predatory catfish in the Cuiabá river basin, northern Pantanal, Brazil.

Fishery is an important economic activity in the Pantanal. Among the regions species, the Pimelodidae catfish stands out as an important part of the annual catch. This study assesses the structure, exploitation and stock management of Hemisorubim platyrhynchos and Sorubim cf. lima, the sixth and seventh largest Pimelodidae of the Pantanal. The analysis is based on fish caught by commercial fishing in the Cuiabá river and landed at the "Ant?nio Moysés Nadaf" Market in the Cuiabá city, Mato Grosso state, Brazil. The findings indicate that commercial fishing activities target several fish cohorts and that usually only individuals above mean length at first maturation are caught. Estimates of the instantaneous mortality coefficient show that the current fishing mortality is low. Simulations of relative yield-per-recruit model demonstrate that the current yield of two species could be greater if the fishery effort were increased, indicating that the stocks are underexploited. However, an increase in current fishery efforts should be viewed with caution, since the stock-recruitment relationship for the species is unknown. The results indicate that the current harvest of two species in the Cuiabá River Basin is sustainable.     

Length‐weight relationships of 26 fish species from the streams of the upper section of the Paraguay River basin (Mato Grosso,Brazil)

This study provides data on the length‐weight relationships (LWR) for 26 tropical fish species collected in August 2010 (dry season) and May 2011 (wet season), and represents the first LWR references for these species in the small tributary streams flowing into the major rivers of the Pantanal Matogrossense, Brazil.     

Automatic greenhouse insect pest detection and recognition based on a cascaded deep learning classification method

Inspection of insect sticky paper traps is an essential task for an effective integrated pest management (IPM) programme. However, identification and counting of the insect pests stuck on the traps is a very cumbersome task. Therefore, an efficient approach is needed to alleviate the problem and to provide timely information on insect pests. In this research, an automatic method for the multi-class recognition of small-size greenhouse insect pests on sticky paper trap images acquired by wireless imaging devices is proposed. The developed algorithm features a cascaded approach that uses a convolutional neural network (CNN) object detector and CNN image classifiers, separately. The object detector was trained for detecting objects in an image, and a CNN classifier was applied to further filter out non-insect objects from the detected objects in the first stage. The obtained insect objects were then further classified into flies (Diptera: Drosophilidae), gnats (Diptera: Sciaridae), thrips (Thysanoptera: Thripidae) and whiteflies (Hemiptera: Aleyrodidae), using a multi-class CNN classifier in the second stage. Advantages of this approach include flexibility in adding more classes to the multi-class insect classifier and sample control strategies to improve classification performance. The algorithm was developed and tested for images taken by multiple wireless imaging devices installed in several greenhouses under natural and variable lighting environments. Based on the testing results from long-term experiments in greenhouses, it was found that the algorithm could achieve average F₁-scores of 0.92 and 0.90 and mean counting accuracies of 0.91 and 0.90, as tested on a separate 6-month image data set and on an image data set from a different greenhouse, respectively. The proposed method in this research resolves important problems for the automated recognition of insect pests and provides instantaneous information of insect pest occurrences in greenhouses, which offers vast potential for developing more efficient IPM strategies in agriculture.     

Small-sized fish as possible seed dispersers: disclosing novel fish and plant species interactions in the Pantanal wetland

ABSTRACT

Seed dispersion is a critical process to vegetation dynamics which is particularly complex in mosaic-like landscapes as the Pantanal wetlands. We analyzed some aspects of ichthyochory in the Pantanal flooded grasslands. The fish were collected and had their whole intestinal contents removed and submitted to suitable germination conditions for 179 days. We collected 661 specimens belonging to 40 species. Feces from Astyanax lacustris, Poptela paraguayensis, Moenkhausia sanctafilomena and Moenkhausia bonita produced 15 seedlings of herbaceous plant of the species Ludwigia inclinata, Ludwigia leptocarpa and a non-identified monocotyledon. A total of 30.4 g of fecal mass was recovered with a rate of 0.49 seedling.gram^?1. There is a positive relation between the number of seedlings obtained and the amount of fish fecal mass (g). Currently, Pantanal landscape dynamics is under rising pressure of field compartmentalization due to man-made roads which restrains fish transit.