Deep learning and computer vision will transform entomology.

Most animal species on Earth are insects, and recent reports suggest that their abundance is in drastic decline. Although these reports come from a wide range of insect taxa and regions, the evidence to assess the extent of the phenomenon is sparse. Insect populations are challenging to study, and most monitoring methods are labor intensive and inefficient. Advances in computer vision and deep learning provide potential new solutions to this global challenge. Cameras and other sensors can effectively, continuously, and noninvasively perform entomological observations throughout diurnal and seasonal cycles. The physical appearance of specimens can also be captured by automated imaging in the laboratory. When trained on these data, deep learning models can provide estimates of insect abundance, biomass, and diversity. Further, deep learning models can quantify variation in phenotypic traits, behavior, and interactions. Here, we connect recent developments in deep learning and computer vision to the urgent demand for more cost-efficient monitoring of insects and other invertebrates. We present examples of sensor-based monitoring of insects. We show how deep learning tools can be applied to exceptionally large datasets to derive ecological information and discuss the challenges that lie ahead for the implementation of such solutions in entomology. We identify four focal areas, which will facilitate this transformation: 1) validation of image-based taxonomic identification; 2) generation of sufficient training data; 3) development of public, curated reference databases; and 4) solutions to integrate deep learning and molecular tools.

Development of underground detection system using a metal detector and aluminum tag for, Copris ochus (Coleoptera: Scarabaeidae).

Tracking of soil-dwelling insects poses greater challenges compared to aboveground-dwelling animals in terrestrial systems. A metal detector system consisting of a commercially available detector and aluminum tags was developed for detecting dung beetle, Copris ochus Motschulsky (Coleoptera: Scarabaeidae). First, detection efficacy of the system was evaluated by varying volumes of aluminum tags attached on a plastic model of the insect and also by varying angles. Then, detection efficacy was evaluated by varying depths of aluminum-tagged models under soil in 2 vegetation types. Finally, the effects of tag attachment on C. ochus adults were assessed for survivorship, burrowing depth, and horizontal movement. Generally, an increase in tag volume resulted in greater detection distance in semi-field conditions. Maximum detection distance of aluminum tag increased up to 17 cm below soil surface as the tag size (0.5 × 1.0 cm [width × length]) and thickness (16 layers) were maximized, resulting in a tag weight of 31.4 mg, comprising ca. 9% of average weight of C. ochus adult. Furthermore, the detection efficacy did not vary among angles except for 90°. In the field, metal detectors successfully detected 5 aluminum-tagged models in 20 × 10 m (W × L) arena within 10 min with detection rates ≥85% for up to depth of 10 cm and 45%-60% at depth of 20 cm. Finally, aluminum tagging did not significantly affect survivorship and behaviors of C. ochus. Our study indicates the potential of metal detector system for tracking C. ochus under soil.

Development of Tenebrio molitor (Coleoptera: Tenebrionidae) on Poultry Litter-Based Diets: Effect on Chemical Composition of Larvae.

In order to investigate a low-cost and sustainable food source, the present study evaluated the use of poultry litter for rearing Tenebrio molitor Linnaeus, 1758 (Coleoptera: Tenebrionidae). The experiment was performed with five diets containing increasing levels of poultry litter (0, 25, 50, 75, and 100%) replacing the control diet and five replicates with 50 larvae per sample unit. Larval growth and development were evaluated and the chemical compositions of diet and T. molitor larvae were determined. Larval development and reproduction efficiency of T. molitor were similar in all treatments. The sole use of poultry litter to feed T. molitor reduced the crude protein of flour by only 8%. Including 50% or more poultry litter in the standard diet is the best-suited formulation for larvae production and incorporation of minerals in the larvae. Mealworm can be grown successfully on diets composed by poultry litter, the diet did not affect survival, growth, and development; however, studies spanning several insect generations should be performed to determine the effects of diet composition on adult fecundity. The knowledge acquired using poultry litter to feed T. molitor will be useful to carry out new research, in addition to evidencing the possibility of low-cost mass rearing of these larvae.

A Review of Insect Monitoring Approaches with Special Reference to Radar Techniques.

Drastic declines in insect populations are a vital concern worldwide. Despite widespread insect monitoring, the significant gaps in the literature must be addressed. Future monitoring techniques must be systematic and global. Advanced technologies and computer solutions are needed. We provide here a review of relevant works to show the high potential for solving the aforementioned problems. Major historical and modern methods of insect monitoring are considered. All major radar solutions are carefully reviewed. Insect monitoring with radar is a well established technique, but it is still a fast-growing topic. The paper provides an updated classification of insect radar sets. Three main groups of insect radar solutions are distinguished: scanning, vertical-looking, and harmonic. Pulsed radar sets are utilized for all three groups, while frequency-modulated continuous-wave (FMCW) systems are applied only for vertical-looking and harmonic insect radar solutions. This work proves the high potential of radar entomology based on the growing research interest, along with the emerging novel setups, compact devices, and data processing approaches. The review exposes promising insect monitoring solutions using compact radar instruments. The proposed compact and resource-effective setups can be very beneficial for systematic insect monitoring.

Saving time maintaining reliability: a new method for quantification of Tetranychus urticae damage in Arabidopsis whole rosettes.

BACKGROUND: The model species Tetranychus urticae produces important plant injury and economic losses in the field. The current accepted method for the quantification of the spider mite damage in Arabidopsis whole rosettes is time consuming and entails a bottleneck for large-scale studies such as mutant screening or quantitative genetic analyses. Here, we describe an improved version of the existing method by designing an automatic protocol. The accuracy, precision, reproducibility and concordance of the new enhanced approach are validated in two Arabidopsis accessions with opposite damage phenotypes. Results are compared to the currently available manual method. RESULTS: Image acquisition experiments revealed that the automatic settings plus 10 values of brightness and the black background are the optimal conditions for a specific recognition of spider mite damage by software programs. Among the different tested methods, the Ilastik-Fiji tandem based on machine learning was the best procedure able to quantify the damage maintaining the differential range of damage between accessions. In addition, the Ilastik-Fiji tandem method showed the lowest variability within a set of conditions and the highest stability under different lighting or background surroundings. Bland-Altman concordance results pointed out a negative value for Ilastik-Fiji, which implies a minor estimation of the damage when compared to the manual standard method. CONCLUSIONS: The novel approach using Ilastik and Fiji programs entails a great improvement for the quantification of the specific spider mite damage in Arabidopsis whole rosettes. The automation of the proposed method based on interactive machine learning eliminates the subjectivity and inter-rater-variability of the previous manual protocol. Besides, this method offers a robust tool for time saving and to avoid the damage overestimation observed with other methods.

Assessment of color response and activity rhythms of the invasive black planthopper Ricania speculum (Walker, 1851) using sticky traps.

To be effective, management strategies of invasive alien species cannot ignore their spatiotemporal behavior particularly those exerting serious damages to human activities. The black planthopper Ricania speculum is an Asian insect that has been reported as an alien invasive species in Italy, where it threatens local plant diversity, including important crops. In our work, we analyzed the activity rhythms of this species through circular statistics and the efficiency of chromotropic traps to capture adult individuals. Captures were carried out in central Italy, where the black planthopper is showing a remarkable range expansion, after its first discovery in 2009. We observed that the species was mainly crepuscular, with a high intersexual activity overlap. Activity rhythms changed between July-August and September-October, with changing heliophany, but peaked at sunset and were the lowest in the second half of the night and early morning. The insects were mostly caught by green traps, particularly in September, which is the period of egg-laying inside the leaves; conversely, orange ones were avoided, and yellow ones captured proportionally to their local availability. Strategies for controlling this species should consider concentrating trapping effort during the activity peak, using green sticky traps to enhance the capture success of each trap, with the lowest impact over non-target species.

A Guide and Toolbox to Replicability and Open Science in Entomology.

The ability to replicate scientific experiments is a cornerstone of the scientific method. Sharing ideas, workflows, data, and protocols facilitates testing the generalizability of results, increases the speed that science progresses, and enhances quality control of published work. Fields of science such as medicine, the social sciences, and the physical sciences have embraced practices designed to increase replicability. Granting agencies, for example, may require data management plans and journals may require data and code availability statements along with the deposition of data and code in publicly available repositories. While many tools commonly used in replicable workflows such as distributed version control systems (e.g., 'git') or script programming languages for data cleaning and analysis may have a steep learning curve, their adoption can increase individual efficiency and facilitate collaborations both within entomology and across disciplines. The open science movement is developing within the discipline of entomology, but practitioners of these concepts or those desiring to work more collaboratively across disciplines may be unsure where or how to embrace these initiatives. This article is meant to introduce some of the tools entomologists can incorporate into their workflows to increase the replicability and openness of their work. We describe these tools and others, recommend additional resources for learning more about these tools, and discuss the benefits to both individuals and the scientific community and potential drawbacks associated with implementing a replicable workflow.

'Drone-Netting' for Sampling Live Insects.

Drones have become valuable tools for biodiversity studies by providing aerial photographs; however, for most entomological studies, images, in particular those taken remotely, are usually insufficient; rather sampling of specimens is required. We equipped a cheap off-the-shelf drone with a net bag, flew it over the ground, sweeping the vegetation, and sampled adult and larval insects as well as spiders. 'Drone-netting' proved to be a versatile method for general insect sampling, particularly in inaccessible terrains. It is time- and cost-effective, minimally invasive, and adaptable for many research tasks in entomofaunistics; it shows a degree of representativeness similar to hand-netting, and caught specimens stay alive and can be released if not needed.

Sampling Efficacy and Survival Rates of Labarrus pseudolividus (Coleoptera: Scarabaeidae) and Onthophagus taurus (Coleoptera: Scarabaeidae) Using Flotation and Sieve-Separation Methodology.

Understanding collection methodologies and their limitations are essential when targeting specific arthropods for use in habitat restoration, conservation, laboratory colony formation, or when holistically representing local populations using ecological surveys. For dung beetles, the most popular collection methodology is baited traps, followed by light traps and unbaited flight-intercept traps during diversity surveys. A less common collection method, flotation, is assumed to be laborious and messy, and so only a handful of papers exist on its refinement and strengths. Our purpose was threefold: First, we tested the recovery and survival rates of Labarrus (=Aphodius) pseudolividus (Balthasar) and Onthophagus taurus (Schreber) when floating beetle-seeded dung pats to determine potential collection and safety issues. We collected 72.4 and 78% of the seeded L. pseudolividus and O. taurus, respectively, with >95% survival rating. Second, we developed a flotation-sieving technique that enables users to rapidly collect and passively sort dung beetles with less time and effort. Specifically, we often collected 50-100 g of wild dung beetles within a couple of hours of gathering dung and sorted them in a couple more by allowing dung beetles to sort themselves by size within a series of sieves; Third, we reviewed flotation-based advantages and disadvantages in comparison to other methodologies.

Using Visual and Digital Imagery to Quantify Horn Fly (Diptera: Muscidae) Densities.

The horn fly, Haematobia irritans L. (Diptera: Muscidae), is a persistent pest of cattle globally. A threshold of 200 flies per animal is considered the standard management goal; however, determining when that threshold has been exceeded is difficult using visual estimates that tend to overestimate the actual fly densities and are, at best, subjective. As a result, a more reliable and durable method of determining horn fly densities is needed. Here, we describe the methods commonly used to quantify horn fly densities including visual estimates and digital photography, and provide examples of quantification software and the prospect for computer automation methods.

Methods for Surveying Stable Fly Populations.

Stable flies are among the most important pests of livestock throughout much of the world. Their painful bites induce costly behavioral and physiological stress responses and reduce productivity. Stable flies are anthropogenic and their population dynamics vary depending on agricultural and animal husbandry practices. Standardized sampling methods are needed to better identify the factors controlling stable fly populations, test novel control technologies, and determine optimal management strategies. The current study reviewed methods used for a long-term study of stable fly population dynamics in the central Great Plains. An additional study compared the relative size of flies sampled from the general population with that of flies sampled emerging from substrates associated with livestock production. Flies developing in livestock associated substrates are significantly larger than those in the general population indicating that other types of developmental sites are contributing significant numbers of flies to the general population. Because efforts to identify those sites have yet to be successful, we speculate that they may be sites with low densities of developing stable flies, but covering large areas such as croplands and grasslands. The stable fly surveillance methods discussed can be used and further improved for monitoring stable fly populations for research and management programs.

Novel Experimental Methods for the Investigation of Hermetia illucens (Diptera: Stratiomyidae) Larvae.

Large-scale insect rearing for food and feed production can be improved by understanding diet digestion and host-microbe interactions. To examine these processes in black soldier fly (Hermetia illucens L.; Diptera: Stratiomyidae) larvae, two protocols were developed. Protocol 1 describes a method to produce viable, sterile black soldier fly larvae and a gentle method for diet sterilization. Sterile black soldier fly larvae can be used to study the diverse role of microbes in larval development. Nutrient requirements of sterile black soldier fly larvae are met only through diet. Viable sterile black soldier fly larvae were consistently generated using a four-step treatment with alternating immersions of eggs for 2 min each in ethanol (70%) and sodium hypochlorite (0.6%), over two cycles. A nonthermal method of diet sterilization, namely high-energy electron beam (HEEB) treatment, was introduced. Subsequently, growth of sterile black soldier fly larvae was observed on the HEEB-treated diets (40, 60, and 40% of replicates with poultry feed, liver pie, and an artificial diet, respectively) but not on autoclaved diets. In Protocol 2, we propose a novel method to collect frass from individual larvae. We then measured the metabolites in frass, using high-pressure liquid chromatography. Results on metabolites confirmed the influence of digestion. For instance, succinate increased from 1 to 2 and 7 µmol/g sample from diet to gut homogenate and frass, respectively. The collection method is a promising tool to estimate the diet and nutrient requirements of black soldier fly larvae, thus increasing the performance and reliability of black soldier fly larvae rearing. We discuss in detail the possible applications and limitations of our methods in black soldier fly larvae research.

Expedited assessment of terrestrial arthropod diversity by coupling Malaise traps with DNA barcoding 1.

Monitoring changes in terrestrial arthropod communities over space and time requires a dramatic increase in the speed and accuracy of processing samples that cannot be achieved with morphological approaches. The combination of DNA barcoding and Malaise traps allows expedited, comprehensive inventories of species abundance whose cost will rapidly decline as high-throughput sequencing technologies advance. Aside from detailing protocols from specimen sorting to data release, this paper describes their use in a survey of arthropod diversity in a national park that examined 21 194 specimens representing 2255 species. These protocols can support arthropod monitoring programs at regional, national, and continental scales.

Mosquito electrocuting traps for directly measuring biting rates and host-preferences of Anopheles arabiensis and Anopheles funestus outdoors.

BACKGROUND: Mosquito biting rates and host preferences are crucial determinants of human exposure to vector-borne diseases and the impact of vector control measures. The human landing catch (HLC) is a gold standard method for measuring human exposure to bites, but presents risks to participants by requiring some exposure to mosquito vectors. Mosquito electrocuting traps (METs) represent an exposure-free alternative to HLCs for measuring human exposure to malaria vectors. However, original MET prototypes were too small for measuring whole-body biting rates on humans or large animals like cattle. Here a much larger MET capable of encompassing humans or cattle was designed, and its performance was evaluated relative to both the original small MET and HLC and for quantifying malaria vector host preferences. METHODS: Human landing catch, small human-baited METs (MET-SH), and large METs baited with either a human (MET-LH) or calves (MET-LC) were simultaneously used to capture wild malaria vectors outdoors in rural southern Tanzania. The four capture methods were compared in a Latin-square design over 20 nights. Malaria vector host preferences were estimated through comparison of the number of mosquitoes caught by large METs baited with either humans or cattle. RESULTS: The MET-LH caught more than twice as many Anopheles arabiensis than either the MET-SH or HLC. It also caught higher number of Anopheles funestus sensu lato (s.l.) compared to the MET-SH or HLC. Similar numbers of An. funestus sensu stricto (s.s.) were caught in MET-LH and MET-SH collections. Catches of An. arabiensis with human or cattle-baited large METs were similar, indicating no clear preference for either host. In contrast, An. funestus s.s. exhibited a strong, but incomplete preference for humans. CONCLUSIONS: METs are a sensitive, practical tool for assessing mosquito biting rates and host preferences, and represent a safer alternative to the HLC. Additionally these findings suggest the HLC underestimate whole-body human exposure. MET collections indicated the An. funestus s.s. population in this setting had a higher than expected attack rate on cattle, potentially making eliminating of this species more difficult with human-targetted control measures. Supplementary vector control tools targetted at livestock may be required to effectively tackle this species.

An assessment of mosquito collection techniques for xenomonitoring of anopheline-transmitted Lymphatic Filariasis in Ghana.

Monitoring vectors is relevant to ascertain transmission of lymphatic filariasis (LF). This may require the best sampling method that can capture high numbers of specific species to give indication of transmission. Gravid anophelines are good indicators for assessing transmission due to close contact with humans through blood meals. This study compared the efficiency of an Anopheles gravid trap (AGT) with other mosquito collection methods including the box and the Centres for Disease Control and Prevention gravid, light, exit and BioGent-sentinel traps, indoor resting collection (IRC) and pyrethrum spray catches across two endemic regions of Ghana. The AGT showed high trapping efficiency by collecting the highest mean number of anophelines per night in the Western (4.6) and Northern (7.3) regions compared with the outdoor collection methods. Additionally, IRC was similarly efficient in the Northern region (8.9) where vectors exhibit a high degree of endophily. AGT also showed good trapping potential for collecting Anopheles melas which is usually difficult to catch with existing methods. Screening of mosquitoes for infection showed a 0.80-3.01% Wuchereria bancrofti and 2.15-3.27% Plasmodium spp. in Anopheles gambiae. The AGT has shown to be appropriate for surveying Anopheles populations and can be useful for xenomonitoring for both LF and malaria.

Active optical sensor assessment of spider mite damage on greenhouse beans and cotton.

The two-spotted spider mite, Tetranychus urticae Koch, is an important pest of cotton in mid-southern USA and causes yield reduction and deprivation in fiber fitness. Cotton and pinto beans grown in the greenhouse were infested with spider mites at the three-leaf and trifoliate stages, respectively. Spider mite damage on cotton and bean canopies expressed as normalized difference vegetation index indicative of changes in plant health was measured for 27 consecutive days. Plant health decreased incrementally for cotton until day 21 when complete destruction occurred. Thereafter, regrowth reversed decline in plant health. On spider mite treated beans, plant vigor plateaued until day 11 when plant health declined incrementally. Results indicate that pinto beans were better suited as a host plant than cotton for rearing T. urticae in the laboratory.

A comparison of trapping techniques (Coleoptera: Carabidae, Buprestidae, Cerambycidae, and Curculionoidea excluding Scolytinae).

Beetles (Coleoptera) are a charismatic group of insects targeted by collectors and often used in biodiversity surveys. As part of a larger project, we surveyed a small (4 hectare) plot in the Boston Mountains of Arkansas using 70 traps of 12 trap types and Berlese-Tullgren extraction of leaf litter and identified all Buprestidae, Carabidae, Cerambycidae, and Curculionoidea (Anthribidae, Attelabidae, Brachyceridae, Brentidae, and Curculionidae excluding Scolytinae) to species. This resulted in the collection of 7,973 specimens representing 242 species arranged in 8 families. In a previous publication, we reported new state records and the number of specimens collected per species. In this publication, we used these data to determine the most effective collection method for four beetle groups: Carabidae, Cerambycidae, Curculionoidea (excluding Scolytinae), and Buprestidae. We found that the combination of pitfall and Malaise traps was most effective for Carabidae, Cerambycidae, and Curculionoidea, but that the combination of Malaise and green Lindgren funnel traps was most effective at collecting Buprestidae. Species accumulation curves did not become asymptotic and extrapolated rarefaction curves did not become asymptotic until 350-1,000 samples, suggesting that much more effort is required to completely inventory even a small site. Additionally, seasonal activity is presented for each species and the similarity and overlap between collecting dates and seasons is discussed for each family.

Infection of Immature Ixodes scapularis (Acari: Ixodidae) by Membrane Feeding.

A reduction in the use of animals in infectious disease research is desirable for animal welfare as well as for simplification and standardization of experiments. An artificial silicone-based membrane-feeding system was adapted for complete engorgement of adult and nymphal Ixodes scapularis Say (Acari: Ixodidae), and for infecting nymphs with pathogenic, tick-borne bacteria. Six wild-type and genetically transformed strains of four species of bacteria were inoculated into sterile bovine blood and fed to ticks. Pathogens were consistently detected in replete nymphs by polymerase chain reaction. Adult ticks that ingested bacteria as nymphs were evaluated for transstadial transmission. Borrelia burgdorferi and Ehrlichia muris-like agent showed high rates of transstadial transmission to adult ticks, whereas Anaplasma phagocytophilum and Rickettsia monacensis demonstrated low rates of transstadial transmission/maintenance. Artificial membrane feeding can be used to routinely maintain nymphal and adult I. scapularis, and infect nymphs with tick-borne pathogens.

The Technical and Performance Characteristics of a Low-Cost, Simply Constructed, Black Light Moth Trap.

The universal mercury vapor black light trap is an effective device used for collecting moth specimens in a wide variety of habitats; yet, they can present challenges for researchers. The mercury vapor trap is often powered by a heavy automotive battery making it difficult to conduct extensive surveys in remote regions. The mercury vapor trap also carries a considerable financial cost per trap unit, making trapping challenging with low research budgets. Here, we describe the development and trapping properties of a lighter, simply constructed, and less expensive trap. The LED funnel trap consists of a funnel, soda bottles with plastic vanes, and is powered by rechargeable 9-V batteries. Two strips of low-wavelength LEDs are used as attractants. We tested the trapping parameters of this trap design compared to a standard mercury vapor trap over 10 trap nights in a suburban woodlot in the summer of 2015. The mercury vapor trap caught significantly more moth individuals than the LED trap (average of 78 vs 40 moths per trap night; P < 0.05), and significantly more species than the LED trap (23 vs 15 per trap night; P < 0.05); the mercury vapor trap caught a total of 104 macromoth species over the duration of the study, compared to a total of 87 by the LED trap. Despite the lower yields, the low cost of the LED trap (<$30 ea.) makes it superior to the mercury vapor trap in cost-acquisition per moth species and per moth individual trapped. The LED trap may be a viable alternative to the standard mercury vapor trap, facilitating insect trapping in more diverse settings.

Development and evaluation of mosquito-electrocuting traps as alternatives to the human landing catch technique for sampling host-seeking malaria vectors.

BACKGROUND: The human landing catch (HLC) is the gold standard method for sampling host-seeking malaria vectors. However, the HLC is ethically questionable because it requires exposure of humans to potentially infectious mosquito bites. METHODS: Two exposure-free methods for sampling host-seeking mosquitoes were evaluated using electrocuting surfaces as potential replacements for HLC: (1) a previously evaluated, commercially available electrocuting grid (CA-EG) designed for killing flies, and (2) a custom-made mosquito electrocuting trap (MET) designed to kill African malaria vectors. The MET and the CA-EG were evaluated relative to the HLC in a Latin Square experiment conducted in the Kilombero Valley, Tanzania. The sampling consistency of the traps across the night and at varying mosquito densities was investigated. Estimates of the proportion of mosquitoes caught indoors (P(i)), proportion of human exposure occurring indoors (π(i)), and proportion of mosquitoes caught when most people are likely to be indoors (P(fl)) were compared for all traps. RESULTS: Whereas the CA-EG performed poorly (<10% of catch of HLC), sampling efficiency of the MET for sampling Anopheles funestus s.l. was indistinguishable from HLC indoors and outdoors. For Anopheles gambiae s.l., sampling sensitivity of MET was 20.9% (95% CI 10.3-42.2) indoors and 58.5% (95% CI 32.2-106.2) outdoors relative to HLC. There was no evidence of density-dependent sampling by the MET or CA-EG. Similar estimates of P(i) were obtained for An. gambiae s.l. and An. funestus s.l. from all trapping methods. The proportion of mosquitoes caught when people are usually indoors (P(fl)) was underestimated by the CA-EG and MET for An. gambiae s.l., but similar to the HLC for An. funestus. Estimates of the proportion of human exposure occurring indoors (π(i)) obtained from the CA-EG and MET were similar to the HLC for An. gambiae s.l., but overestimated for An. funestus. CONCLUSIONS: The MET showed promise as an outdoor sampling tool for malaria vectors where it achieved >50% sampling sensitivity relative to the HLC. The CA-EG had poor sampling sensitivity outdoors and inside. With further modification, the MET could provide an efficient and safer alternative to the HLC for the surveillance of mosquito vectors outdoors.