Surveying moths using light traps: effects of weather and time of year

Towards a standardized framework for AI-assisted, image-based monitoring of nocturnal insects

Automated sensors have potential to standardize and expand the monitoring of insects across the globe. As one of the most scalable and fastest developing sensor technologies, we describe a framework for automated, image-based monitoring of nocturnal insects-from sensor development and field deployment to workflows for data processing and publishing. Sensors comprise a light to attract insects, a camera for collecting images and a computer for scheduling, data storage and processing. Metadata is important to describe sampling schedules that balance the capture of relevant ecological information against power and data storage limitations. Large data volumes of images from automated systems necessitate scalable and effective data processing. We describe computer vision approaches for the detection, tracking and classification of insects, including models built from existing aggregations of labelled insect images. Data from automated camera systems necessitate approaches that account for inherent biases. We advocate models that explicitly correct for bias in species occurrence or abundance estimates resulting from the imperfect detection of species or individuals present during sampling occasions. We propose ten priorities towards a step-change in automated monitoring of nocturnal insects, a vital task in the face of rapid biodiversity loss from global threats. This article is part of the theme issue 'Towards a toolkit for global insect biodiversity monitoring'.

Object Detection of Small Insects in Time-Lapse Camera Recordings

As pollinators, insects play a crucial role in ecosystem management and world food production. However, insect populations are declining, necessitating efficient insect monitoring methods. Existing methods analyze video or time-lapse images of insects in nature, but analysis is challenging as insects are small objects in complex and dynamic natural vegetation scenes. In this work, we provide a dataset of primarily honeybees visiting three different plant species during two months of the summer. The dataset consists of 107,387 annotated time-lapse images from multiple cameras, including 9423 annotated insects. We present a method for detecting insects in time-lapse RGB images, which consists of a two-step process. Firstly, the time-lapse RGB images are preprocessed to enhance insects in the images. This motion-informed enhancement technique uses motion and colors to enhance insects in images. Secondly, the enhanced images are subsequently fed into a convolutional neural network (CNN) object detector. The method improves on the deep learning object detectors You Only Look Once (YOLO) and faster region-based CNN (Faster R-CNN). Using motion-informed enhancement, the YOLO detector improves the average micro F1-score from 0.49 to 0.71, and the Faster R-CNN detector improves the average micro F1-score from 0.32 to 0.56. Our dataset and proposed method provide a step forward for automating the time-lapse camera monitoring of flying insects.

Anthropogenic Influence on Moth Populations: A Comparative Study in Southern Sweden

As moths are vital components of ecosystems and serve as important bioindicators, understanding the dynamics of their communities and the factors influencing these dynamics, such as anthropogenic impacts, is crucial to understand the ecological processes. Our study focuses on two provinces in southern Sweden, Västergötland and Småland, where we used province records from 1974 to 2019 in combination with light traps (in 2020) to record the presence and abundance of moth species, subsequently assessing species traits to determine potential associations with their presence in anthropogenically modified landscapes. This study design provides a unique opportunity to assess temporal changes in moth communities and their responses to shifts in environmental conditions, including anthropogenic impacts. Across the Västergötland and Småland provinces in Sweden, we recorded 776 moth taxa belonging to fourteen different taxonomic families of mainly Macroheterocera. We captured 44% and 28% of the total moth species known from these provinces in our traps in Borås (Västergötland) and Kalmar (Småland), respectively. In 2020, the species richness and abundance were higher in Borås than in Kalmar, while the Shannon and Simpson diversity indices revealed a higher species diversity in Kalmar. Between 1974 and 2019, the colonisation rates of the provinces increased faster in Småland. Ninety-three species were found to have colonised these provinces since 1974, showing that species richness increased over the study period. We reveal significant associations between the probability of a species being present in the traps and distinct traits compared to a provincial species pool. Traits over-represented in the traps included species with a high variation in colour patterns, generalist habitat preferences, extended flight periods, lower host plant specificity, and overwintering primarily as eggs. Our findings underscore the ongoing ecological filtering that favours certain species-specific traits. This study sheds light on the roles of climate change and anthropogenic impacts in shaping moth biodiversity, offers key insights into the ecological processes involved, and can guide future conservation efforts.

The Contribution of Singletons and Doubletons Captured Using Weak Light Heath Traps for the Analysis of the Macroheteroceran Assemblages in Forest Biotopes

In nearly every ecological community, most species are represented by a few individuals, and most individuals come from a few of the most common species. Singletons (one individual sampled) and doubletons (two individuals sampled) are very common in moth community studies. In some reports, these specimens are excluded from the analysis once they are considered a consequence of under-sampling or of contamination with tourist species that are just passing through. Throughout 12 nights in 2018 and 12 nights in 2019, two Heath traps, one with an 8 W ultraviolet lamp and the other with a 15 W actinic lamp, were positioned approximately 50 m apart at nine sites of four different biotopes in a mosaic forest ecosystem in the Narew National Park (NE Poland). We were able to differentiate moth assemblages according to the forest biotopes under study and by the year of research. With our results, it becomes more evident that singletons and doubletons sampled using weak light Heath traps should be included in the ecological analysis of Macroheteroceran moth assemblages, and our research strongly suggests that they are an important and consistent element of such a sampling method. We also demonstrate that weak light Heath traps are suitable for building an inventory scheme of moth assemblages in small forest areas and that singletons and doubletons can be crucial elements in long-term monitoring systems.

Towards the fully automated monitoring of ecological communities

High-resolution monitoring is fundamental to understand ecosystems dynamics in an era of global change and biodiversity declines. While real-time and automated monitoring of abiotic components has been possible for some time, monitoring biotic components-for example, individual behaviours and traits, and species abundance and distribution-is far more challenging. Recent technological advancements offer potential solutions to achieve this through: (i) increasingly affordable high-throughput recording hardware, which can collect rich multidimensional data, and (ii) increasingly accessible artificial intelligence approaches, which can extract ecological knowledge from large datasets. However, automating the monitoring of facets of ecological communities via such technologies has primarily been achieved at low spatiotemporal resolutions within limited steps of the monitoring workflow. Here, we review existing technologies for data recording and processing that enable automated monitoring of ecological communities. We then present novel frameworks that combine such technologies, forming fully automated pipelines to detect, track, classify and count multiple species, and record behavioural and morphological traits, at resolutions which have previously been impossible to achieve. Based on these rapidly developing technologies, we illustrate a solution to one of the greatest challenges in ecology: the ability to rapidly generate high-resolution, multidimensional and standardised data across complex ecologies.

Weather Sensitivity of Sugar Bait Trapping of Nocturnal Moths: A Case Study from Northern Europe

Assemblages of insects need to be quantitatively sampled in the context of various research questions. Light trapping is the most widely used method for sampling nocturnal Lepidoptera. Attracting moths to sugar baits offers a viable alternative. However, this method is rarely used in professional research despite its popularity among amateur lepidopterists. As the activity of insects is strongly dependent on ambient conditions, the sensitivity of any trapping method to weather parameters needs to be known for the quantitative interpretation of trapping results. In the present paper, we report data on the weather dependence of moth catches obtained by automatic bait traps. The study was performed in Estonia, representing the European hemiboreal forest zone. Portable weather stations set up next to each of the traps were used for collecting weather data. Both abundance and diversity of the moths in the catches depended strongly positively on temperature and negatively on air humidity. Diversity was also negatively correlated with air pressure and positively with the change in pressure during the night. The results show that in situ recording of weather parameters in connection to insect trapping provides useful insights for the study of insect behaviour and the interpretation of the results of monitoring projects.

A preliminary checklist of moths (Lepidoptera: Heterocera) from Gangajalghati, Bankura, West Bengal, India

The present study was conducted at Gangajalghati, a village near the forest of Bankura district from West Bengal that has a tropical wet and dry climate where moth diversity has not been explored before. The village was surveyed between January 2016 and December 2018. The present study has recorded a total of 1,328 individual moths belonging to 13 families, 31 subfamilies, 80 genera, and 90 species. Three species—Condylorrhiza diniasalis (Walker, 1859), Argyrocosma inductaria (Guenée, 1858), and Oraesia emarginata (Fabricius, 1794)—are reported for the first time from West Bengal and Eublemma roseonivea (Walker, 1863) shows its westernmost distribution in West Bengal, India. It was earlier reported from India (Assam), China, Taiwan, the Philippines, Malaya, and Borneo.

Light pollution impairs urban nocturnal pollinators but less so in areas with high tree cover

The increase in artificial light at night (ALAN) is widely considered as a major driver for the worldwide decline of nocturnal pollinators such as moths. However, the relationship between light and trees as 'islands of shade' within urban areas has not yet been fully understood. Here, we studied (1) the effects of three landscape variables, i.e. sources of ALAN (mercury vapour/LED street lamps; overall light pollution), impervious surfaces (e.g. roads, parking lots and buildings), and tree cover on species richness and abundance of two major macro-moth families (Noctuidae and Geometridae) and (2) the potential mitigating effect of trees on macro-moths attracted to ALAN. We undertook a landscape-scale study on 22 open green areas along an urban-rural gradient within Berlin, Germany, using light traps to collect moths. Macro-moths were identified to species level and GLMMs applied with the three landscape variables at different scales (100 m, 500 m and 1000 m). We found a significant negative effect of mercury vapour street lamps on macro-moth species richness, while impervious surfaces showed significant negative effects on abundance (total and Geometridae). We further found significant positive effects of tree cover density on species richness and abundance (total and Geometridae). Effects of tree cover, however, were mostly driven by one site. LED lamps showed no predictive effects. A negative effect of ALAN (MV lamps and overall light) on macro-moths was most prominent in areas with low tree coverage, indicating a mitigating effect of trees on ALAN. We conclude that mercury vapour street lamps should be replaced by ecologically more neutral ALAN, and that in lit and open areas trees could be planted to mitigate the negative effect of ALAN on nocturnal pollinators. In addition, sources of ALAN should be carefully managed, using movement detection technology and other means to ensure that light is only produced when necessary.

Forest microclimates and climate change: Importance, drivers and future research agenda

Forest microclimates contrast strongly with the climate outside forests. To fully understand and better predict how forests' biodiversity and functions relate to climate and climate change, microclimates need to be integrated into ecological research. Despite the potentially broad impact of microclimates on the response of forest ecosystems to global change, our understanding of how microclimates within and below tree canopies modulate biotic responses to global change at the species, community and ecosystem level is still limited. Here, we review how spatial and temporal variation in forest microclimates result from an interplay of forest features, local water balance, topography and landscape composition. We first stress and exemplify the importance of considering forest microclimates to understand variation in biodiversity and ecosystem functions across forest landscapes. Next, we explain how macroclimate warming (of the free atmosphere) can affect microclimates, and vice versa, via interactions with land-use changes across different biomes. Finally, we perform a priority ranking of future research avenues at the interface of microclimate ecology and global change biology, with a specific focus on three key themes: (1) disentangling the abiotic and biotic drivers and feedbacks of forest microclimates; (2) global and regional mapping and predictions of forest microclimates; and (3) the impacts of microclimate on forest biodiversity and ecosystem functioning in the face of climate change. The availability of microclimatic data will significantly increase in the coming decades, characterizing climate variability at unprecedented spatial and temporal scales relevant to biological processes in forests. This will revolutionize our understanding of the dynamics, drivers and implications of forest microclimates on biodiversity and ecological functions, and the impacts of global changes. In order to support the sustainable use of forests and to secure their biodiversity and ecosystem services for future generations, microclimates cannot be ignored.

Qualitative and Quantitative Loss of Habitat at Different Spatial Scales Affects Functional Moth Diversity

Land use change has led to large-scale insect decline, threatening ecosystem resilience through reduced functional diversity. Even in nature reserves, losses in insect diversity have been detected. Hereby, changes in local habitat quality and landscape-scale habitat quantity can play a role driving functional diversity toward erosion. Our aim was to analyze how local and landscape-scale factors simultaneously affect functional insect diversity. Therefore, we sampled moths in two Italian coastal forest reserves at 60 sites. Our focus was on functional richness, redundancy and niche occupation, being important for ecosystem resilience, following the insurance framework. Ecological information about 387 species and 14 traits was used to analyze functional diversity. Twenty-five functional groups were recognized and used to estimate niche occupation and redundancy. Fourteen local and 12 landscape-scale factors were measured and condensed by using Principal Components Analysis. The resulting PC-axes served as predictors in linear mixed effects models. Functional richness, redundancy and niche occupation of moths were lower at sites with low habitat quality and quantity, indicating reduced ecosystem resilience. Especially landscape diversity and habitat structure, viz. a humidity-nutrient gradient, but also plant diversity, were promoting functional richness. Landscape fragmentation, indicating increased impermeability for insects, reduced local functional richness, redundancy and niche occupation. Local habitat quality and landscape-wide habitat quantity are both important for maintaining functional insect diversity inside reserves. Therefore, small and isolated nature reserves might fail in preserving biodiversity and ecosystem functions through adverse effects acting from the surrounding landscape structure and configuration.

Standards and Best Practices for Monitoring and Benchmarking Insects

Benchmark studies of insect populations are increasingly relevant and needed amid accelerating concern about insect trends in the Anthropocene. The growing recognition that insect populations may be in decline has given rise to a renewed call for insect population monitoring by scientists, and a desire from the broader public to participate in insect surveys. However, due to the immense diversity of insects and a vast assortment of data collection methods, there is a general lack of standardization in insect monitoring methods, such that a sudden and unplanned expansion of data collection may fail to meet its ecological potential or conservation needs without a coordinated focus on standards and best practices. To begin to address this problem, we provide simple guidelines for maximizing return on proven inventory methods that will provide insect benchmarking data suitable for a variety of ecological responses, including occurrence and distribution, phenology, abundance and biomass, and diversity and species composition. To track these responses, we present seven primary insect sampling methods—malaise trapping, light trapping, pan trapping, pitfall trappings, beating sheets, acoustic monitoring, and active visual surveys—and recommend standards while highlighting examples of model programs. For each method, we discuss key topics such as recommended spatial and temporal scales of sampling, important metadata to track, and degree of replication needed to produce rigorous estimates of ecological responses. We additionally suggest protocols for scalable insect monitoring, from backyards to national parks. Overall, we aim to compile a resource that can be used by diverse individuals and organizations seeking to initiate or improve insect monitoring programs in this era of rapid change.

An Automated Light Trap to Monitor Moths (Lepidoptera) Using Computer Vision-Based Tracking and Deep Learning

Insect monitoring methods are typically very time-consuming and involve substantial investment in species identification following manual trapping in the field. Insect traps are often only serviced weekly, resulting in low temporal resolution of the monitoring data, which hampers the ecological interpretation. This paper presents a portable computer vision system capable of attracting and detecting live insects. More specifically, the paper proposes detection and classification of species by recording images of live individuals attracted to a light trap. An Automated Moth Trap (AMT) with multiple light sources and a camera was designed to attract and monitor live insects during twilight and night hours. A computer vision algorithm referred to as Moth Classification and Counting (MCC), based on deep learning analysis of the captured images, tracked and counted the number of insects and identified moth species. Observations over 48 nights resulted in the capture of more than 250,000 images with an average of 5675 images per night. A customized convolutional neural network was trained on 2000 labeled images of live moths represented by eight different classes, achieving a high validation F1-score of 0.93. The algorithm measured an average classification and tracking F1-score of 0.71 and a tracking detection rate of 0.79. Overall, the proposed computer vision system and algorithm showed promising results as a low-cost solution for non-destructive and automatic monitoring of moths.

Management of flying insects on expressways through an academic-industrial collaboration: evaluation of the effect of light wavelengths and meteorological factors on insect attraction

Insect outbreaks often occur in the absence of natural enemies and in the presence of excess suitable host materials. Outbreaks of gypsy moths are especially problematic in remote areas located in high-latitude regions in Japan because the majority of adults emerge during the short summer season and initiate synchronous mass flight toward artificial lights. The aggregation of moths in public facilities not only is an annoyance to visitors but also permits the establishment of new populations the following year. The aim of this study was to establish a method to reduce the numbers of large moths that are attracted to lights in the rest areas of expressways in Hokkaido based on the results of research on their behavioral ecology and physiology. First, we conducted extensive insect surveys using light traps that emit light at different wavelengths; the traps were set along the expressways in the summers of 2014-2018. The insects attracted to the light were roughly classified into those showing a preference for broadband light wavelengths (from UV-A to green) and short light wavelengths (from UV-A to blue). The former included aquatic insects and winged ants, and the latter included moths and beetles. Next, we analyzed correlations between moth emergence and daily meteorological data. When gypsy moths were abundant during an outbreak, the daily catch of gypsy moths was positively correlated with the highest ambient temperature on the catch day but not with the visibility range, wind speed, or moon phase. In contrast, the daily catch of oak silkmoths did not correlate with any of these parameters. Our results provide guidance for the management of forest insects inhabiting cool-temperate to subarctic regions based on light wavelengths with reference to weather variables.

Effects of Lethal Bronzing Disease, Palm Height, and Temperature on Abundance and Monitoring of Haplaxius crudus

Insect vector feeding preference and behavior play important roles in pathogen transmission, especially for pathogens that solely rely on insect vector transmission. This study aims to examine the effects of the 16SrIV-D phytoplasma, the causal agent of lethal bronzing (LB) disease of palms, on associated auchenorrhynchan insects. The numbers of auchenorrhynchans collected during weekly surveys during a yearlong study using yellow sticky traps were analyzed. The cumulative number of H. crudus was 4.5 times greater on phytoplasma-infected relative to non-infected palms. Other auchenorrhynchans showed no difference between phytoplasma-infected and non-infected palms or were greater on non-infected rather than on infected palms. Furthermore, we examined the effects of LB, palm height, temperature, and the interactive effects of these factors on H. crudus abundance. When the palms were infected with LB, at low temperature, H. crudus was more abundant on shorter than taller palms; however, H. crudus was more abundant on taller than shorter palms at the median and higher temperatures. These results may indicate that H. crudus prefers LB-infected palms over non-infected palms. The interactive effects of LB, palm heights, and temperature further suggest that vector monitoring and disease management should be optimized according to seasonal variation in temperature.

Moth responses to sympathetic hedgerow management in temperate farmland

Hedgerows provide valuable habitats and corridors for many species in farmland, yet a lack of appropriate management may threaten their benefits to biodiversity. Although agri-environment scheme (AES) prescriptions on hedgerow management have the potential to reverse the detrimental effect of over-trimming on wildlife, their effectiveness has rarely been addressed. The aims of the study were to (i) assess moth responses to trimming regimes; and (ii) investigate the influence of the surrounding landscape on moth assemblages. We specifically tested the effectiveness of the trimming regime recommended by the targeted AES that was implemented on farms near greater horseshoe bat (Rhinolophus ferrumequinum) colonies since it represented the most sympathetic hedgerow management option among English AES options. We sampled adult micro- and macro-moths along 64 hedgerows located within 20 English farms using light traps, and classified moths into two guilds reflecting their larval food preferences, namely grass/herb- and shrub/tree-feeders. Our results suggest that reducing trimming has a positive impact on macro-moth species richness as well as on shrub/tree-feeder abundance and species richness. It also benefited four moth species that are significantly declining in Britain. Furthermore, while the proportion of woodland at a large spatial scale (3.0 km radius around the sampling sites) was positively associated with the abundance of macro-moths and grass/herb-feeders, woodland connectivity had a positive effect on the species richness of grass/herb- and shrub/tree-feeders at large and medium (1.5 km radius) scales, respectively. Both the abundance and species richness of macro-moths and the abundance of shrub/tree-feeders were negatively affected by the presence of arable fields adjacent to hedgerows. Overall, these findings reveal the wider biodiversity benefits of targeted AESs focusing on habitat improvement for R. ferrumequinum, and the importance of woodland in the wider landscape. We therefore strongly recommend implementing a multi-scale management approach (i.e. from field to landscape) through the use of adequate AES prescriptions to conserve moths in agricultural landscapes.

Plant diversity enhances moth diversity in an intensive forest management experiment

Intensive forest management (IFM) promises to help satisfy increasing global demand for wood but may come at the cost of local reductions to forest biodiversity. IFM often reduces early seral plant diversity as a result of efforts to eliminate plant competition with crop trees. If diversity is a function of bottom-up drivers, theory predicts that specialists at lower trophic levels (e.g., insect herbivores) should be particularly sensitive to reductions in plant diversity. We conducted a stand-level experiment to test bottom-up controls on moth community structure, as mediated by degrees of forest management intensity. Using a dataset of 12,003 moths representing 316 moth species, moth richness decreased only slightly, if at all, as herbicide intensity increased (P = 0.062); the moderate treatment, which is most commonly applied in the northwestern USA, was estimated to have 4.72 (±2.14 SE, P = 0.039) fewer species than the control. Structural equation modeling revealed strong support for an effect of herbicide on plant abundance, which influenced plant species richness and subsequently moth species richness. Moth species richness was associated with plant species richness and followed a power law function (z = 0.42, P = 0.006), which is surprisingly consistent with a recent large-scale experiment in agricultural systems, and provides support for bottom-up drivers of moth community structure. Moth abundance was not influenced by the direct effects of silvicultural herbicide treatments. Site-level effects and variation in pre-harvest vegetation communities resulted in residual broadleaf and herbaceous vegetation in even the most intensive treatment. Even at low densities, these residual deciduous and herbaceous plants supported higher than expected moth abundance and richness. We conclude that forest management practices that retain early seral vegetation diversity are the most likely to conserve moth communities.

Variable coloration is associated with dampened population fluctuations in noctuid moths

Theory and recent reviews state that greater genetic and phenotypic variation should be beneficial for population abundance and stability. Experimental evaluations of this prediction are rare, of short duration and conducted under controlled environmental settings. The question whether greater diversity in functionally important traits stabilizes populations under more complex ecological conditions in the wild has not been systematically evaluated. Moths are mainly nocturnal, with a large variation in colour patterns among species, and constitute an important food source for many types of organisms. Here, we report the results of a long-term (2003-2013) monitoring study of 115 100 noctuid moths from 246 species. Analysis of time-series data provide rare evidence that species with higher levels of inter-individual variation in colour pattern have higher average abundances and undergo smaller between-year fluctuations compared with species having less variable colour patterns. The signature of interspecific temporal synchronization of abundance fluctuations was weak, suggesting that the dynamics were driven by species-specific biotic interactions rather than by some common, density-independent factor(s). We conclude that individual variation in colour patterns dampens population abundance fluctuations, and suggest that this may partly reflect that colour pattern polymorphism provides protection from visually oriented predators and parasitoids.