Drift Samples Research Articles

Herbicide residue detection in cotton as influenced by time, drift rate, and sampling method

Abstract Herbicide drift to sensitive crops can result in significant injury, yield loss, and even crop destruction. When pesticide drift is reported to the Georgia Department of Agriculture (GDA), tissue samples are collected and analyzed for residues. Seven field studies were conducted in 2020 and 2021 in cooperation with the GDA to evaluate the effect of (1) time interval between simulated drift event and sampling, (2) low-dose herbicide rates, and (3) the sample collection methods on detecting herbicide residues in cotton (Gossypium hirsutum L.) foliage. Simulated drift rates of 2,4-D, dicamba, and imazapyr were applied to non-tolerant cotton in the 8- to 9-leaf stage with plant samples collected at 7 or 21 d after treatment (DAT). During collection, plant sampling consisted of removing entire plants or removing new growth occurring after the 7-leaf stage. Visual cotton injury from 2,4-D reached 43% to 75% at 0.001 and 0.004 kg ae ha−1, respectively; for dicamba, it was 9% to 41% at 0.003 or 0.014 kg ae ha−1, respectively; and for imazapyr, it was 1% to 74% with 0.004 and 0.03 kg ae ha−1 rates, respectively. Yield loss was observed with both rates of 2,4-D (11% to 51%) and with the high rate of imazapyr (52%); dicamba did not influence yield. Herbicide residues were detected in 88%, 88%, and 69% of samples collected from plants treated with 2,4-D, dicamba, and imazapyr, respectively, at 7 DAT compared with 25%, 16%, and 22% when samples were collected at 21 DAT, highlighting the importance of sampling quickly after a drift event. Although the interval between drift event and sampling, drift rate, and sampling method can all influence residue detection for 2,4-D, dicamba, and imazapyr, the factor with the greatest influence is the amount of time between drift and sample collection.

Data Quality-Aware Client Selection in Heterogeneous Federated Learning

Federated Learning (FL) enables decentralized data utilization while maintaining edge user privacy, but it faces challenges due to statistical heterogeneity. Existing approaches address client drift and data heterogeneity issues. However, real-world settings often involve low-quality data with noisy features, such as covariate drift or adversarial samples, which are usually ignored. Noisy samples significantly impact the global model’s accuracy and convergence rate. Assessing data quality and selectively aggregating updates from high-quality clients is crucial, but dynamically perceiving data quality without additional computations or data exchanges is challenging. In this paper, we introduce the FedDQA (Federated learning via Data Quality-Aware) (FedDQA) framework. We discover increased data noise leads to slower loss reduction during local model training. We propose a loss sharpness-based Data-Quality-Awareness (DQA) metric to differentiate between high-quality and low-quality data. Based on the DQA, we design a client selection algorithm that strategically selects participant clients to reduce the negative impact of noisy clients. Experiment results indicate that FedDQA significantly outperforms the baselines. Notably, it achieves up to a 4% increase in global model accuracy and demonstrates faster convergence rates.

Investigating the influence of measurement uncertainty on chlorophyll-a predictions as an indicator of harmful algal blooms in machine learning models

Advancements in data availability, including high frequency, near real-time multiparameter sensors, laboratory analysis, and in-situ and remote observations, have driven the development of machine learning (ML) models for applications such as toxic Harmful Algal Bloom (HABs) monitoring. However, the performance of ML predictions is influenced by both model uncertainties due to inherent model structures and errors associated with input dataset measurements. For example, measurement uncertainty arises from sample collection, sensor drift and laboratory analysis and sample handling errors. While impacts of model uncertainty are commonly addressed using probabilistic approaches, the effect of measurement uncertainty is less studied due to the limited availability of detailed measurement information. This study focuses on assessing the impact of measurement uncertainty on the ML prediction of chlorophyll-a concentration as an index of HABs in a mesotrophic lake. Using randomized subsets of input measured datasets that mimic possible chlorophyll-a concentration distributions, the study built 1000 Random Forest (RF) and Support Vector Regression (SVR) models. An independent measured dataset was used to validate the ensemble models, allowing for model performance evaluation and the creation of prediction intervals to measure the propagated uncertainty. Our findings showed that the model predictions have MAE that ranged between 0.16 μg/l and 5.19 μg/l, and RMSE ranging between 0.20 μg/l and 7.39 μg/l. The highest uncertainty coverage of 0.71 was observed in the RF model without chlorophyll-a sensor values as a predictor. The study found that the training dataset sizes due to the high frequency and manually sampled nature influence how much measurement uncertainty is covered. The results of this study demonstrate how well ML models can capture various HABs patterns when given diverse measurement variables. Our findings will give researchers insightful information on how to lessen the impact of measurement uncertainty when using ML models as decision-support tools for HABs management.

An empirical evaluation of meta residual network for classifying sensor drift samples

With the advancements in materials and pattern recognition algorithms, electronic nose (E-nose) has been continuously developing and finding increasingly widespread applications. However, E-nose is prone to drift during practical usage. Currently, the primary methods for addressing the issue of sensor drift in sample classification are semi-supervised learning and unsupervised learning. The implementation of supervised learning has been difficult in the past due to the requirement for a large number of up-to-date samples to retrain the models. But now, this challenge can be overcome with the help of meta-learning. We applied the model-agnostic meta-learning (MAML) architecture from meta-learning and designed a residual classifier with learnable per-parameter learning rates. Leveraging the ability of ‘learning to learn’, our method do not require a large amount of target domain data and there is no need to repeat extensive computations when transferring to any other target domain compared to other algorithms, resulting it highly deployable. Through the evaluation of multiple experiments, we demonstrate the feasibility of our approach and achieve very high classification accuracy on a publicly available long-term drift dataset compared with other methods.

Taxonomic accuracy and complementarity between bulk and eDNA metabarcoding provides an alternative to morphology for biological assessment of freshwater macroinvertebrates

Determining biological status of freshwater ecosystems is critical for ensuring ecosystem health and maintaining associated services to such ecosystems. Freshwater macroinvertebrates respond predictably to environmental disturbances and are widely used in biomonitoring programs. However, many freshwater species are difficult to capture and sort from debris or substrate and morphological identification is challenging, especially larval stages, damaged specimens, or hyperdiverse groups such as Diptera. The advent of high throughput sequencing technologies has enhanced DNA barcoding tools to automatise species identification for whole communities, as metabarcoding is increasingly used to monitor biodiversity. However, recent comparisons have revealed little congruence between morphological and molecular-based identifications. Using broad range universal primers for DNA barcode marker cox1, we compare community composition captured between morphological and molecular-based approaches from different sources — tissue-based (bulk benthic and bulk drift samples) and environmental DNA (eDNA, filtered water) metabarcoding — for samples collected along a gradient of anthropogenic disturbances. For comparability, metabarcoding taxonomic assignments were filtered by taxa included in the standardised national biological metric IBMWP. At the family level, bulk benthic metabarcoding showed the highest congruence with morphology, and the most abundant taxa were captured by all techniques. Richness captured by morphology and bulk benthic metabarcoding decreased along the gradient, whereas richness recorded by eDNA remained constant and increased downstream when sequencing bulk drift. Estimates of biological metrics were higher using molecular than morphological identification. At species level, diversity captured by bulk benthic samples were higher than the other techniques. Importantly, bulk benthic and eDNA metabarcoding captured different and complementary portions of the community — benthic versus water column, respectively — and their combined use is recommended. While bulk benthic metabarcoding can likely replace morphology using similar benthic biological indices, water eDNA will require new metrics because this technique sequences a different portion of the community.

The prey availability and diet of juvenile Atlantic salmon (Salmo salar L.) in low-productivity rivers in northern Europe.

The availability of resources varies across a species distributional range, and a low-productivity area can make a species more vulnerable. We investigated the invertebrate composition and prey choice of juvenile Atlantic salmon (Salmo salar L.) in low-productivity rivers in northeast Iceland, which is one of the species' most northerly distributions. By sampling benthic and drift invertebrate populations, we found that prey availability was similar within and between rivers. Gut content samples showed that the main prey choice for juvenile S. salar was the Chironomidae. The type of food items consumed varied across different weight groups of S. salar, with smaller juveniles having more diverse diet. S. salar did not have a selection preference for chironomids, which indicates that they were eating the highly available prey in their environment, rather than hunting high biomass items such as terrestrial invertebrates and large Dipterans. Estimates of dietary niche showed that S. salar in these low-productivity rivers relied on consuming what was most readily available, the chironomids, and that they must share resources with other salmonid species. This may be due to the low diversity of freshwater invertebrates (fewer prey options), whereas S. salar in nutrient-rich rivers could rely more on terrestrial invertebrates as an additional subsidy in their diet. In conclusion, with limited prey choices, juvenile S. salar in nutrient-poor rivers, especially in a biogeographically isolated region with low species diversity, may increase in vulnerability and decrease in adaptability to environmental change. Management methods that increase benthic prey abundance and diversity are recommended for conserving the S. salar population in a nutrient-poor river.

Formation of complex discrete samples for electrocardiogram isoline drift extraction devices

In the automatic analysis of an electrocardiosignal, in order to adequately assess its amplitude parameters, it is necessary to ensure that there is no isoline drift in the signal. The direct removal of the isoline drift using a high-pass filter leads to distortion of the shape and amplitude parameters of the electrocardiosignal elements. When extracting the isoline drift based on interpolation of samples on the PQ segment, the frequency of the extracted isoline drift signal cannot even theoretically exceed half the heart rate, but in reality it is an order of magnitude less. The purpose is to develop a device for the formation of complex discrete samples of the electrocardiosignal, ensuring the isolation without loss of informative components of the electrocardiosignal of the isoline drift with a frequency range comparable to the heart rate, in conditions of variability of the sampling frequency due to heart rate variability. The possibility of expanding the frequency range of the isoline drift signal isolated from the electrocardiosignal based on the real-time transformation of isoline drift samples taken at the TP interval into complex discrete samples has been substantiated. When forming complex discrete samples, the variability of the sampling period due to heart rate variability is taken into account. A device for forming complex discrete samples of the isoline drift signal has been proposed, allowing extraction of isoline drift with an extended frequency range without loss the informative components of the electrocardiosignal. The developed device can be used both directly as part of mobile electrocardiographs designed to monitor a person's ECG in the conditions of his daily activities, and in equipment for processing daily electrocardiograms before analyzing their parameters.

Interval Type-2 Fuzzy Neural Network Based on Active Semi-Supervised Learning for Non-Stationary Industrial Processes

Accurate models ensure the efficient and safe operations of industrial processes. However, modeling of industrial processes with non-stationary characteristic is challenging. In this study, an interval type-2 fuzzy neural network (IT2FNN) based on active semi-supervised learning (ASSL-IT2FNN) is proposed for such industrial processes. First, an IT2FNN with adaptive fuzzy membership function (FMF) and hierarchical learning method is utilized to achieve considerable modeling accuracy and efficiency. Second, to better tackle with the non-stationary characteristic, an unsupervised distribution detection method is proposed to identify the occurrence moments of concept drift actively from the perspective of probability and spatial projection. Then, by an active semi-supervised learning method, the concept drift samples with partial labels are used to build a subnetwork, guaranteeing the performance of the constructed IT2FNN models with incomplete data in non-stationary environments. Finally, the proposed ASSL-IT2FNN is verified by benchmark simulations and real industrial data, and the experimental results demonstrate its outperformance. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —Concept drift is an important indicator of a non-stationary environment. It manifests as changes in the distribution of samples over time, which may reduce the accuracy of models constructed based on historical samples. This study aims to alleviate the impact of concept drift on model performance in non-stationary industrial processes while reduce the cost of sample labeling. By using the adaptive FMF and hierarchical learning method, the designed IT2FNN can effectively deal with the inherent nonlinearity of complicated industrial processes. The samples representing concept drift are actively identified by the Kullback-Leibler (KL) divergence and maximum mean difference (MMD), enabling timely detections of concept drift moments in unsupervised situations. The active semi-supervised learning is used to adapt the constructed model to handle the concept drift, wherein a subnetwork is designed based on concept drift samples and similar historical samples.

BenchMFC: A benchmark dataset for trustworthy malware family classification under concept drift

Concept drift poses a critical challenge in deploying machine learning models to mitigate practical malware threats. It refers to the phenomenon that the distribution of test data changes over time, gradually deviating from the original training data and degrading model performance. A promising direction for addressing concept drift is to detect drift samples and then retrain the model. However, this field currently lacks a unified, well-curated, and comprehensive benchmark, which often leads to unfair comparisons and inconclusive outcomes. To improve the evaluation and advance further, this paper presents a new Benchmark dataset for trustworthy Malware Family Classification (BenchMFC), which includes 223 K samples of 526 families that evolve over years. BenchMFC provides clear family, packer, and timestamp tags for each sample, it thus can support research on three types of malware concept drift: 1) unseen families, 2) packed families, and 3) evolved families. To collect unpacked family samples from large-scale candidates, we introduce a novel crowdsourcing malware annotation pipeline, which unifies packing detection and family annotation as a consensus inference problem to prevent costly packing detection. Moreover, we provide two case studies to illustrate the application of BenchMFC in 1) concept drift detection and 2) model retraining. The first case demonstrates the impact of three types of malware concept drift and compares nine notable concept drift detectors. The results show that existing detectors have their own advantages in dealing with different types of malware concept drift, and there is still room for improvement in malware concept drift detection. The second case explores how static feature-based machine learning operates on packed samples when retraining a model. The experiments illustrate that packers do preserve some kind of signals that appear to be “effective” for machine learning models, but the robustness of these signals requires further research. BenchMFC has been released to the community at https://github.com/crowdma/benchmfc.

FeSAD ransomware detection framework with machine learning using adaption to concept drift

This paper proposes FeSAD, a framework that will allow a machine learning classifier to detect evolutionary ransomware. Ransomware is a critical player in the malware space that causes hundreds of millions of dollars of damage globally and evolves quickly. The evolution of ransomware in machine learning classifiers is often calculated as concept drift. Concept drift is dangerous as changes in the behavior of ransomware can easily lead to misclassifications, and misclassification can harm individuals and businesses. Our proposed framework consists of a feature selection layer, drift calibration layer and drift decision layer that allows a machine learning classifier to detect and classify concept drift samples reliably. We evaluate the FeSAD framework in various concept drift scenarios and observe its ability to detect drifting samples effectively. The FeSAD framework is also evaluated on its ability to extend the lifespan of a classifier. The results obtained by this research show that FeSAD can successfully and reliably classify ransomware and benign samples while under concept drift and can extend the time between retraining.

A bold new purpose for an old method: Using invertebrate kick‐netting to improve monitoring of microplastic pollution in running waters

AbstractFreshwaters are impacted by many pollutants. Scientists and resource managers need to reliably detect and monitor these pollutants by employing appropriate sampling techniques to quantify and mitigate their impacts. Emerging freshwater contaminants such as microplastics are difficult to monitor as effective sampling techniques have not been fully developed and lack standardization. Here, we propose a novel method, adapted from stream invertebrate kick‐net sampling, which can be employed for long‐term, standardized microplastic monitoring. We hypothesized that invertebrate kick‐net sampling would be effective at collecting microplastics, capturing higher microplastic concentrations than standard microplastic sampling, due to collecting suspended and benthic microplastics simultaneously. We sampled 28 streams for microplastics using standard drift and sediment microplastic collection methods and invertebrate kick‐netting. As predicted, kick‐netting captured microplastics at higher concentrations than conventional sampling, regardless of whether values were expressed per volume of water (as in drift samples), per kg of sediment or per area sampled (as in benthic samples). Consequently, kick‐net sampling has the potential to be a time‐ and cost‐efficient tool for monitoring microplastic pollution. We recommend the employment of invertebrate kick‐netting as a new, standardized means to investigate and routinely monitor microplastic concentrations worldwide. This would generate a more robust dataset of global freshwater microplastic pollution, making it possible to answer unresolved questions pertaining to changes in microplastic pollution profiles. Standardized, long‐term records of microplastic concentrations in freshwaters would also allow a more accurate assessment of the ecological risks of microplastic pollution. Finally, long‐term microplastic data could be used to inform much‐needed regulatory decisions pertaining to microplastic pollution.

Efficient concept drift handling for batch android malware detection models

The rapidly evolving nature of Android apps poses a significant challenge to static batch machine learning algorithms employed in malware detection systems, as they quickly become obsolete. Despite this challenge, the existing literature pays limited attention to addressing this issue, with many advanced Android malware detection approaches, such as Drebin, DroidDet and MaMaDroid, relying on static models. In this work, we show how retraining techniques are able to maintain detector capabilities over time. Particularly, we analyze the effect of two aspects in the efficiency and performance of the detectors: (1) the frequency with which the models are retrained, and (2) the data used for retraining. In the first experiment, we compare periodic retraining with a more advanced concept drift detection method that triggers retraining only when necessary. In the second experiment, we analyze sampling methods to reduce the amount of data used to retrain models. Specifically, we compare fixed sized windows of recent data and state-of-the-art active learning methods that select those apps that help keep the training dataset small but diverse. Our experiments show that concept drift detection and sample selection mechanisms result in very efficient retraining strategies which can be successfully used to maintain the performance of the static Android malware state-of-the-art detectors in changing environments.

Comparing passive and active macroinvertebrate sampling gear efficacy during biomonitoring in Southern Appalachian mountain streams

AbstractAquatic invertebrates are important components of any aquatic ecosystem and are frequently monitored to determine the ecological integrity of those systems. Various gear types are used to collect aquatic macroinvertebrates in wadeable streams. We analyzed commonly used active and passive gear types, including kick seines, Hess samplers, drift nets, and Hester–Dendy samplers, across a watershed in the Great Smoky Mountains National Park to determine their efficacy under various habitat conditions in wadeable Southern Appalachian mountain streams. Hester–Dendy samplers were significantly different from other gear types for most univariate and multivariate analyses. Kick and Hess samples were most similar in multivariate comparisons of macroinvertebrate assemblages. Drift samples were similar in most univariate analyses but were distinct from other gear types when analyzed with multivariate techniques. Our results indicate that comprehensive assessments of macroinvertebrate assemblages require multiple gear types, various analytical approaches, and consistent taxonomic resolution. The complexity of the fluvial system in question should be thoroughly assessed prior to formulating a macroinvertebrate biomonitoring protocol to better ensure robust and meaningful results.

Cluster-Guided Contrastive Graph Clustering Network

Benefiting from the intrinsic supervision information exploitation capability, contrastive learning has achieved promising performance in the field of deep graph clustering recently. However, we observe that two drawbacks of the positive and negative sample construction mechanisms limit the performance of existing algorithms from further improvement. 1) The quality of positive samples heavily depends on the carefully designed data augmentations, while inappropriate data augmentations would easily lead to the semantic drift and indiscriminative positive samples. 2) The constructed negative samples are not reliable for ignoring important clustering information. To solve these problems, we propose a Cluster-guided Contrastive deep Graph Clustering network (CCGC) by mining the intrinsic supervision information in the high-confidence clustering results. Specifically, instead of conducting complex node or edge perturbation, we construct two views of the graph by designing special Siamese encoders whose weights are not shared between the sibling sub-networks. Then, guided by the high-confidence clustering information, we carefully select and construct the positive samples from the same high-confidence cluster in two views. Moreover, to construct semantic meaningful negative sample pairs, we regard the centers of different high-confidence clusters as negative samples, thus improving the discriminative capability and reliability of the constructed sample pairs. Lastly, we design an objective function to pull close the samples from the same cluster while pushing away those from other clusters by maximizing and minimizing the cross-view cosine similarity between positive and negative samples. Extensive experimental results on six datasets demonstrate the effectiveness of CCGC compared with the existing state-of-the-art algorithms. The code of CCGC is available at https://github.com/xihongyang1999/CCGC on Github.

Possible seasonal and diurnal modulation of Gammarus pulex (Crustacea, Amphipoda) drift by microsporidian parasites

In lotic freshwater ecosystems, the drift or downstream movement of animals (e.g., macroinvertebrates) constitutes a key dispersal pathway, thus shaping ecological and evolutionary patterns. There is evidence that macroinvertebrate drift may be modulated by parasites. However, most studies on parasite modulation of host drifting behavior have focused on acanthocephalans, whereas other parasites, such as microsporidians, have been largely neglected. This study provides new insight into possible seasonal and diurnal modulation of amphipod (Crustacea: Gammaridae) drift by microsporidian parasites. Three 72 h drift experiments were deployed in a German lowland stream in October 2021, April, and July 2022. The prevalence and composition of ten microsporidian parasites in Gammarus pulex clade E varied seasonally, diurnally, and between drifting and stationary specimens of G. pulex. Prevalence was generally higher in drifting amphipods than in stationary ones, mainly due to differences in host size. However, for two parasites, the prevalence in drift samples was highest during daytime suggesting changes in host phototaxis likely related to the parasite’s mode of transmission and site of infection. Alterations in drifting behavior may have important implications for G. pulex population dynamics and microsporidians’ dispersal. The underlying mechanisms are more complex than previously thought.

The Riverine Organism Drift Imager: A new technology to study organism drift in rivers and streams

Abstract Drift or downstream dispersal is a fundamental process in the life cycle of many riverine organisms. In the face of rapidly declining freshwater biodiversity, there is a need to enhance our capacity to study the drift of riverine organisms, by overcoming the limitations of traditional labour‐intensive sampling methods that result in data of low temporal and spatial resolution. To address this need, we developed a new technology, the Riverine Organism Drift Imager (RODI), which combines in situ imaging with machine‐learning classification. This technique expands on the traditional methodology by replacing the collection cup of a drift net with a camera system that continuously images riverine organisms as they drift through the device. After being imaged, organisms are released into the environment unharmed. A machine‐learning classifier is used after field sampling to identify drifting organisms. Therefore, RODI provides a non‐invasive sampling method that can quantify organism drift at unprecedented temporal resolution. Multiple deployments have served to validate the performance of the technology in the field. In its current implementation, images are captured continuously for 1.5 h at 50 frames per second. We demonstrate that the quality of the resulting images enables a convolutional neural network classifier to identify organisms to the family level. The weighted F1 score, a metric for the performance of the classifier, was 94%, based on training and testing on a field‐collected dataset consisting of 4598 images of 285 organisms belonging to seven classes (one species, five families and one order). In conclusion, this work provides a proof of concept, demonstrating the viability of the deployment of RODI as an automated, in situ organism drift sampler. This novel approach offers the possibility to advance our fundamental understanding of the drift of riverine organisms and how this is affected by human impacts in natural streams while, at the same time, can serve as a cost‐effective tool for biodiversity monitoring.

Clustering-based Active Learning Classification towards Data Stream

Many practical applications, such as social media and monitoring system, will constantly generate streaming data, which has problems of instability, lack of labels and multiclass imbalance. In order to solve these problems, a cluster-based active learning method is proposed to achieve data stream classification. Firstly, a label query strategy combining marginal threshold matrix is proposed, which selects difficult to classify or potential concept drift samples for marking, to solve the problem of high cost label and unbalanced data. Secondly, dynamic maintenance of a group of micro clusters, by adjusting the weight of micro clusters in the model, correctly reflects the current data distribution, and finally, uses the buffer to store new micro clusters to participate in the update of the model, to adapt to the new data environment. Experimental results on three real data sets and three synthetic data sets show that compared with the classical data stream classification algorithm, it is less affected by concept drift and has higher classification accuracy than the online semi-supervised learning algorithm ADSM. The average accuracy of the six datasets increased by 5.56%, 2.32%, 1.77%, 1.83%, 3.78%, and 2.04%, respectively. The model processes data streams online and improves classification performance with less memory consumption.

Comparison of Weather Acquisition Periods Influencing a Statistical Model of Aerial Pesticide Drift

Off-target drift of crop protection materials from aerial spraying can be detrimental to sensitive crops, beneficial insects, and the environment. So, it is very important to accurately characterize weather effects for accurate recommendations on drift mitigation. Wind is the single-most important weather factor influencing localized off-target drift of crop protection materials. In drift sampling experiments, it is difficult to accurately characterize wind speed and direction at a drift sampling location, owing to the natural variability of spray movement on the way to the sampling target. Although it is difficult or impossible to exactly track wind movement to a target, much information can be gained by altering the way wind speed and tracking is characterized from field experiments and analyzed using statistical models of spray drift. In this study two methods of characterizing weather were compared to see how they affect results from a statistical model of downwind spray drift using field data. Use of a method that implemented weather averages over the length of a spray run resulted in model-based estimates for spray tracer concentration that compared well with field data averages. Model results using this method showed only a slight sensitivity to changes in wind speed, and this difference was more pronounced further downwind. The degree of this effect was consistent with field results. Another method that used single weather values obtained at the beginning of each run resulted in an unexpected inverse relationship of residue concentration with respect to increases in wind speed by sensitivity analysis and would thus not be recommended for use in a statistical model of downwind spray drift. This study could provide a guideline for general agricultural aviation analysis and unmanned aerial vehicle spray application studies.

Declines in prey production during the collapse of a tailwater Rainbow Trout population are associated with changing reservoir conditions

AbstractObjectiveUnderstanding how energy moves through food webs and limits productivity at various trophic levels is a central question in aquatic ecology and can provide insight into drivers of fish population dynamics since many fish populations are food limited. In this study, we seek to better understand what factors drove a decline of &gt;85% in the number of Rainbow TroutOncorhynchus mykiss found in the tailwater portion of the Colorado River below Glen Canyon Dam during 2012–2016.MethodsWe estimate the production of dominant prey using data from previously published studies of Rainbow Trout abundance and growth alongside drift and diet samples. We test how prey production correlates to both proximate (e.g., nutrients) and distal (e.g., limnological conditions in the upriver reservoir) drivers.ResultResults suggest that gross consumption of invertebrate prey by the Rainbow Trout population declined from an annual mean of 423 to 69 kg/d. Daily production rates of dominant prey in aggregate declined from a high of 0.173 to 0.018 g·m−2·d−1. Chironomids accounted for 70% of the decline in prey production. Foraging efficiency by Rainbow Trout (range, 0.99–0.67) was high across the range of prey production rates. After the Rainbow Trout population had declined by ~90%, prey consumption saturated at higher rates of prey production and the gross quantity of daily drift exported from the reach increased from 8.9 to 12.7 kg/d.ConclusionRainbow Trout population dynamics are largely influenced by changes in prey production, which is itself driven by soluble reactive phosphorus (SRP) concentrations in the reservoir. The SRP model predicted that prey production would increase by 32 kg/d (SE, 9) for each 1 μg/L increase in SRP. These concentrations were indirectly influenced by reservoir hydrology and biogeochemistry, linkages that may extend far beyond the confines of this tailwater fishery and into the downstream reaches of the Grand Canyon's Colorado River ecosystem.

Domain Adaptation on Asymmetric Drift Data for an Electronic Nose

Domain adaptation, a type of transfer learning, has been theoretically adopted to the electronic nose (EN) drift problem. Current academic achievements in this field are inspired by symmetric data, requiring identical sample categories between source and target domains. Commonly, the source domain comprises prepared EN data with intact ground truth; the target domain contains EN drift data to be recognized without class labels. However, in practical cases, the category number of collected drift data is always less than that of prepared data. Hence, performance degradation would be occurred using traditional domain adaptation. To address this problem, we modified and fused dictionary learning, canonical correlation analysis, and locality preserving projection for the asymmetry domain data. Accordingly, selective first-order, second-order alignments, and topology preservation have been realized via sparse and generalized regulation. Specifically, we presented an associated iterative solution process to gain the projected subspace. Then, we utilized two drift datasets to create several asymmetric EN drift scenarios for performance validation. The experimental results showed the highest recognition accuracies achieved by the proposed method on asymmetric drift data in most scenarios. The proposed methodology was conducted with higher accuracy and stability in drift sample recognition than the other referenced drift compensation methods. It is a suitable choice for EN drift compensation on asymmetric data.