Extensive data engineering to the rescue: building a multi-species katydid detector from unbalanced, atypical training datasets.

Passive acoustic monitoring (PAM) is a powerful tool for studying ecosystems. However, its effective application in tropical environments, particularly for insects, poses distinct challenges. Neotropical katydids produce complex species-specific calls, spanning mere milliseconds to seconds and spread across broad audible and ultrasonic frequencies. However, subtle differences in inter-pulse intervals or central frequencies are often the only discriminatory traits. These extremities, coupled with low source levels and susceptibility to masking by ambient noise, challenge species identification in PAM recordings. This study aimed to develop a deep learning-based solution to automate the recognition of 31 katydid species of interest in a biodiverse Panamanian forest with over 80 katydid species. Besides the innate challenges, our efforts were also encumbered by a limited and imbalanced initial training dataset comprising domain-mismatched recordings. To overcome these, we applied rigorous data engineering, improving input variance through controlled playback re-recordings and by employing physics-based data augmentation techniques, and tuning signal-processing, model and training parameters to produce a custom well-fit solution. Methods developed here are incorporated into Koogu, an open-source Python-based toolbox for developing deep learning-based bioacoustic analysis solutions. The parametric implementations offer a valuable resource, enhancing the capabilities of PAM for studying insects in tropical ecosystems. This article is part of the theme issue 'Towards a toolkit for global insect biodiversity monitoring'.

Multi-year soundscape recordings and automated call detection reveals varied impact of moonlight on calling activity of neotropical forest katydids.

Night-time light can have profound ecological effects, even when the source is natural moonlight. The impacts of light can, however, vary substantially by taxon, habitat and geographical region. We used a custom machine learning model built with the Python package Koogu to investigate the in situ effects of moonlight on the calling activity of neotropical forest katydids over multiple years. We prioritised species with calls that were commonly detected in human annotated data, enabling us to evaluate model performance. We focused on eight species of katydids that the model identified with high precision (generally greater than 0.90) and moderate-to-high recall (minimum 0.35), ensuring that detections were generally correct and that many calls were detected. These results suggest that moonlight has modest effects on the amount of calling, with the magnitude and direction of effect varying by species: half of the species showed positive effects of light and half showed negative. These findings emphasize the importance of understanding natural history for anticipating how biological communities respond to moonlight. The methods applied in this project highlight the emerging opportunities for evaluating large quantities of data with machine learning models to address ecological questions over space and time. This article is part of the theme issue 'Towards a toolkit for global insect biodiversity monitoring'.

Levels of Airborne Sound And Substrate-borne Vibration Calling Are Negatively Related Across Neotropical False-leaf Katydids.

Animals often signal in multiple sensory modalities to attract mates, but the level of signaling investment in each modality can differ dramatically between individuals and across species. When functionally overlapping signals are produced in different modalities, their relative use can be influenced by many factors, including differences in signal active space, energetic costs, and predation risk. Characterizing differences in total signal investment across time can shed light on these factors, but requires long focal recordings of signal production. Neotropical pseudophylline katydids produce mate advertisement signals as airborne sound and substrate-borne vibration. Airborne calls, produced via stridulation, are extremely short, high-frequency, and longer-range signals. Conversely, substrate-borne calls produced via abdominal tremulation are longer, low-frequency, relatively more energetically costly, and shorter-range signals. To examine patterns of stridulation and tremulation across species and test hypotheses about the drivers of signal use in each modality, we recorded multimodal signaling activity over 24 hours for males from ten pseudophylline species from a single Panamanian community. We also collected data on demographic and morphological species characteristics, and acoustic features of airborne calls, such as bandwidth, peak frequency, and duration. Finally, we generated a molecular phylogeny for these species and used phylogenetic generalized least squares models to test for relationships between variables while controlling for evolutionary relationships. We found a negative relationship between sound and vibration calling, indicating that substrate-borne vibrational signaling may compensate for reduced airborne signaling in these species. Sound call bandwidth and the proportion of males collected at lights, a proxy for the amount of male movement, also explained a significant amount of variation in sound calling across species, indicating that the overall relationship between the two types of calling signals may be mediated by the specific characteristics of the signals as well as other species traits.

Changes in Cough Airflow and Acoustics After Injection Laryngoplasty.

OBJECTIVE/HYPOTHESIS: We explored the following hypotheses in a cohort of patients undergoing injection laryngoplasty: (1) glottic insufficiency affects voluntary cough airflow dynamics and restoring glottic competence may improve parameters of cough strength, (2) cough strength can be inferred from cough acoustic signal, and (3) glottic competence changes cough sounds and correlates with spectrogram morphology. STUDY TYPE/DESIGN: Prospective interventional study. METHODS: Subjects with glottic insufficiency secondary to unilateral vocal fold paresis, paralysis, or atrophy, and scheduled for injection laryngoplasty completed an instrumental assessment of voluntary cough airflow using a pneumotachometer and a protocolized voluntary cough sound recording. A Wilcoxon signed-rank test was used to compare the differences between pre- and post-injection laryngoplasty in airflow and acoustic measures. A Spearman rank-order correlation was used to evaluate the association between airflow and acoustic cough measures. RESULTS: Twenty-five patients (13F:12M, mean age 68.8) completed voluntary cough airflow measurements and 22 completed cough sound recordings. Following injection laryngoplasty, patients had a statistically significant decreased peak expiratory flow rise time (PEFRT) (mean change: -0.03 s, SD: 0.06, p = 0.04) and increased cough volume acceleration (mean change: 13.1 L/s2 , SD: 33.9, p = 0.03), suggesting improved cough effectiveness. Correlation of cough acoustic measures with airflow measures showed a weak relationship between PEFRT and acoustic energy (coefficient: -0.31, p = 0.04) and peak power density (coefficient: -0.35, p = 0.02). CONCLUSIONS: Our study thus indicates that injection laryngoplasty may help avert aspiration in patients with glottic insufficiency by improving cough effectiveness and that improved cough airflow measures may be tracked with cough sounds. LEVEL OF EVIDENCE: 3 Laryngoscope, 133:S1-S14, 2023.

Patterns of Herbivory in Neotropical Forest Katydids as Revealed by DNA Barcoding of Digestive Tract Contents.

Many well-studied animal species use conspicuous, repetitive signals that attract both mates and predators. Orthopterans (crickets, katydids, and grasshoppers) are renowned for their acoustic signals. In Neotropical forests, however, many katydid species produce extremely short signals, totaling only a few seconds of sound per night, likely in response to predation by acoustically orienting predators. The rare signals of these katydid species raises the question of how they find conspecific mates in a structurally complex rainforest. While acoustic mechanisms, such as duetting, likely facilitate mate finding, we test the hypothesis that mate finding is further facilitated by colocalization on particular host plant species. DNA barcoding allows us to identify recently consumed plants from katydid stomach contents. We use DNA barcoding to test the prediction that katydids of the same species will have closely related plant species in their stomach. We do not find evidence for dietary specialization. Instead, katydids consumed a wide mix of plants within and across the flowering plants (27 species in 22 genera, 16 families, and 12 orders) with particular representation in the orders Fabales and Laurales. Some evidence indicates that katydids may gather on plants during a narrow window of rapid leaf out, but additional investigations are required to determine whether katydid mate finding is facilitated by gathering at transient food resources.

Analytical approaches for evaluating passive acoustic monitoring data: A case study of avian vocalizations.

The interface between field biology and technology is energizing the collection of vast quantities of environmental data. Passive acoustic monitoring, the use of unattended recording devices to capture environmental sound, is an example where technological advances have facilitated an influx of data that routinely exceeds the capacity for analysis. Computational advances, particularly the integration of machine learning approaches, will support data extraction efforts. However, the analysis and interpretation of these data will require parallel growth in conceptual and technical approaches for data analysis. Here, we use a large hand-annotated dataset to showcase analysis approaches that will become increasingly useful as datasets grow and data extraction can be partially automated.We propose and demonstrate seven technical approaches for analyzing bioacoustic data. These include the following: (1) generating species lists and descriptions of vocal variation, (2) assessing how abiotic factors (e.g., rain and wind) impact vocalization rates, (3) testing for differences in community vocalization activity across sites and habitat types, (4) quantifying the phenology of vocal activity, (5) testing for spatiotemporal correlations in vocalizations within species, (6) among species, and (7) using rarefaction analysis to quantify diversity and optimize bioacoustic sampling.To demonstrate these approaches, we sampled in 2016 and 2018 and used hand annotations of 129,866 bird vocalizations from two forests in New Hampshire, USA, including sites in the Hubbard Brook Experiment Forest where bioacoustic data could be integrated with more than 50 years of observer-based avian studies. Acoustic monitoring revealed differences in community patterns in vocalization activity between forests of different ages, as well as between nearby similar watersheds. Of numerous environmental variables that were evaluated, background noise was most clearly related to vocalization rates. The songbird community included one cluster of species where vocalization rates declined as ambient noise increased and another cluster where vocalization rates declined over the nesting season. In some common species, the number of vocalizations produced per day was correlated at scales of up to 15 km. Rarefaction analyses showed that adding sampling sites increased species detections more than adding sampling days.Although our analyses used hand-annotated data, the methods will extend readily to large-scale automated detection of vocalization events. Such data are likely to become increasingly available as autonomous recording units become more advanced, affordable, and power efficient. Passive acoustic monitoring with human or automated identification at the species level offers growing potential to complement observer-based studies of avian ecology.

Individual contributions to group chorus dynamics influence access to mating opportunities in wood frogs.

A limitation in bioacoustic studies has been the inability to differentiate individual sonic contributions from group-level dynamics. We present a novel application of acoustic camera technology to investigate how individual wood frogs' calls influence chorus properties, and how variation influences mating opportunities. We recorded mating calls and used playback trials to gauge preference for different chorus types in the laboratory. Males and females preferred chorus playbacks with low variance in dominant frequency. Females preferred choruses with low mean peak frequency. Field studies revealed more egg masses laid in ponds where males chorused with low variance in dominant frequency. We also noted a trend towards more egg masses laid in ponds where males called with low mean frequency. Nearest-neighbour distances influenced call timing (neighbours called in succession) and distances increased with variance in chorus frequency. Results highlight the potential fitness implications of individual-level contributions to a bioacoustic signal produced by groups.

Daily Signaling Rate and the Duration of Sound per Signal are Negatively Related in Neotropical Forest Katydids.

Researchers have long examined the structure of animal advertisement signals, but comparatively little is known about how often these signals are repeated and what factors predict variation in signaling rate across species. Here, we focus on acoustic advertisement signals to test the hypothesis that calling males experience a tradeoff between investment in the duration or complexity of individual calls and investment in signaling over long time periods. This hypothesis predicts that the number of signals that a male produces per 24 h will negatively correlate with (1) the duration of sound that is produced in each call (the sum of all pulses) and (2) the number of sound pulses per call. To test this hypothesis, we measured call parameters and the number of calls produced per 24 h in 16 species of sympatric phaneropterine katydids from the Panamanian rainforest. This assemblage also provided us with the opportunity to test a second taxonomically specific hypothesis about signaling rates in taxa such as phaneropterine katydids that transition from advertisement calls to mating duets to facilitate mate localization. To establish duets, male phaneropterine katydids call and females produce a short acoustic reply. These duets facilitate searching by males, females, or both sexes, depending on the species. We test the hypothesis that males invest either in calling or in searching for females. This hypothesis predicts a negative relationship between how often males signal over 24 h and how much males move across the landscape relative to females. For the first hypothesis, there was a strong negative relationship between the number of signals and the duration of sound that is produced in each signal, but we find no relationship between the number of signals produced per 24 h and the number of pulses per signal. This result suggests the presence of cross-taxa tradeoffs that limit signal production and duration, but not the structure of individual signals. These tradeoffs could be driven by energetic limitations, predation pressure, signal efficacy, or other signaling costs. For the second hypothesis, we find a negative relationship between the number of signals produced per day and proportion of the light trap catch that is male, likely reflecting males investing either in calling or in searching. These cross-taxa relationships point to the presence of pervasive trade-offs that fundamentally shape the spatial and temporal dynamics of communication.

Sheep in wolves' clothing: prey rely on proactive defences when predator and non-predator cues are similar.

Predation produces intense selection and a diversity of defences. Reactive defences are triggered by predator cues, whereas proactive defences are always in effect. We assess whether prey rely on proactive defences when predator cues do not correlate well with predation risk. Many bats use echolocation to hunt insects, and many insects have evolved to hear bats. However, in species-rich environments like Neotropical forests, bats have extremely diverse foraging strategies, and the presence of echolocation corresponds only weakly to the presence of predators. We assess whether katydids that live in habitats with many non-dangerous bat species stop calling when exposed to echolocation. For 11 species of katydids, we quantified behavioural and neural responses to predator cues, and katydid signalling activity over 24 h periods. Despite having the sensory capacity to detect predators, many Neotropical forest katydids continued calling in the presence of predator cues, displaying proactive defences instead (short, infrequent calls totalling less than 2 cumulative seconds of sound per 24 h). Neotropical katydid signalling illustrates a fascinating case where trophic interactions are probably mediated by a third group: bats with alternative foraging strategies (e.g. frugivory). Although these co-occurring bats are not trophically connected, their mere presence disrupts the correlation between cue and predation risk.

Applying and refining DNA analysis to determine the identity of plant material extracted from the digestive tracts of katydids.

BACKGROUND: Feeding habits are central to animal ecology, but it is often difficult to characterize the diet of organisms that are arboreal, nocturnal, rare, or highly mobile. Genetic analysis of gut contents is a promising approach for expanding our understanding of animal feeding habits. Here, we adapt a laboratory protocol for extracting and sequencing plant material from gut contents and apply it to Neotropical forest katydids (Orthoptera: Tettigoniidae) on Barro Colorado Island (BCI) in Panama. METHODS: Our approach uses three chloroplast primer sets that were previously developed to identify vegetation on BCI. We describe the utility and success rate of each primer set. We then test whether there is a significant difference in the amplification and sequencing success of gut contents based on the size or sex of the katydid, the time of day that it was caught, and the color of the extracted gut contents. RESULTS: We find that there is a significant difference in sequencing success as a function of gut color. When extracts were yellow, green, or colorless the likelihood of successfully amplifying DNA ranged from ~30-60%. When gut extracts were red, orange, or brown, amplification success was exceptionally low (0-8%). Amplification success was also higher for smaller katydids and tended to be more successful in katydids that were captured earlier in the night. Strength of the amplified product was indicative of the likelihood of sequencing success, with strong bands having a high likelihood of success. By anticipating which samples are most likely to succeed, we provide information useful for estimating the number of katydids that need to be collected and minimizing the costs of purifying, amplifying, and sequencing samples that are unlikely to succeed. This approach makes it possible to understand the herbivory patterns of these trophically important katydids and can be applied more broadly to understand the diet of other tropical herbivores.

Theory Meets Empiry: A Citation Network Analysis.

According to a recent survey, ecologists and evolutionary biologists feel that theoretical and empirical research should coexist in a tight feedback loop but believe that the two domains actually interact very little. We evaluate this perception using a citation network analysis for two data sets, representing the literature on sexual selection and speciation. Overall, 54%-60% of citations come from a paper's own category, whereas 17%-23% are citations across categories. These cross-citations tend to focus on highly cited papers, and we observe a positive correlation between the numbers of citations a study receives within and across categories. We find evidence that reviews can function as integrators between the two literatures, argue that theoretical models are analogous to specific empirical study systems, and complement our analyses by studying a cocitation network. We conclude that theoretical and empirical research are more tightly connected than generally thought but that avenues exist to further increase this integration.

Mechanisms of Assortative Mating in Speciation with Gene Flow: Connecting Theory and Empirical Research.

The large body of theory on speciation with gene flow has brought to light fundamental differences in the effects of two types of mating rules on speciation: preference/trait rules, in which divergence in both (female) preferences and (male) mating traits is necessary for assortment, and matching rules, in which individuals mate with like individuals on the basis of the presence of traits or alleles that they have in common. These rules can emerge from a variety of behavioral or other mechanisms in ways that are not always obvious. We discuss the theoretical properties of both types of rules and explain why speciation is generally thought to be more likely under matching rather than preference/trait rules. We furthermore discuss whether specific assortative mating mechanisms fall under a preference/trait or matching rule, present empirical evidence for these mechanisms, and propose empirical tests that could distinguish between them. The synthesis of the theoretical literature on these assortative mating rules with empirical studies of the mechanisms by which they act can provide important insights into the occurrence of speciation with gene flow. Finally, by providing a clear framework we hope to inspire greater alignment in the ways that both theoreticians and empiricists study mating rules and how these rules affect speciation through maintaining or eroding barriers to gene flow among closely related species or populations.

Beyond temperature coupling: Effects of temperature on ectotherm signaling and mate choice and the implications for communication in multispecies assemblages.

Many organisms share communication channels, generating complex signaling environments that increase the risk of signal interference. Variation in abiotic conditions, such as temperature, may further exacerbate signal interference, particularly in ectotherms. We tested the effects of temperature on the pulse rate of male signals in a community of Oecanthus tree crickets, and for one focal species we also assessed its effect on female pulse rate preferences and motivation to seek mates. We confirm prior findings of temperature-dependent signals that result in increasing signal similarity at lower temperatures. Temperature also affected several aspects of female preferences: The preferred pulse rate value was temperature dependent, and nearly perfectly coupled with signal pulse rate; the range of pulse rate values that females found attractive also increased with temperature. By contrast, the motivation of females to perform phonotaxis was unaffected by temperature. Thus, at lower temperatures the signals of closely related species were more similar and females more discriminating. However, because signal similarity increased more strongly than female discrimination, signal interference and the likelihood of mismating may increase as temperatures drop. We suggest that a community approach will be useful for understanding the role of environmental variability in the evolution of communication systems.

The role of ecology in speciation by sexual selection: a systematic empirical review.

Theoretical and empirical research indicates that sexual selection interacts with the ecological context in which mate choice occurs, suggesting that sexual and natural selection act together during the evolution of premating reproductive isolation. However, the relative importance of natural and sexual selection to speciation remains poorly understood. Here, we applied a recent conceptual framework for examining interactions between mate choice divergence and ecological context to a review of the empirical literature on speciation by sexual selection. This framework defines two types of interactions between mate choice and ecology: internal interactions, wherein natural and sexual selection jointly influence divergence in sexual signal traits and preferences, and external interactions, wherein sexual selection alone acts on traits and preferences but ecological context shapes the transmission efficacy of sexual signals. The objectives of this synthesis were 3-fold: to summarize the traits, ecological factors, taxa, and geographic contexts involved in studies of mate choice divergence; to analyze patterns of association between these variables; and to identify the most common types of interactions between mate choice and ecological factors. Our analysis revealed that certain traits are consistently associated with certain ecological factors. Moreover, among studies that examined a divergent sexually selected trait and an ecological factor, internal interactions were more common than external interactions. Trait-preference associations may thus frequently be subject to both sexual and natural selection in cases of divergent mate choice. Our results highlight the importance of interactions between sexual selection and ecology in mate choice divergence and suggest areas for future research.

Community composition affects the shape of mate response functions.

The evolution of mate preferences can be critical for the evolution of reproductive isolation and speciation. Heterospecific interference may carry substantial fitness costs and result in preferences where females are most responsive to the mean conspecific trait with low response to traits that differ from this value. However, when male traits are unbounded by heterospecifics, there may not be selection against females that respond to extreme trait values in the unbounded direction. To test how heterospecifics affected the shape of female response functions, I presented female Oecanthus tree crickets with synthetic calls representing a range of male calls, then measured female phonotaxis to construct response functions. The species with the fastest pulse rates in the community consistently responded to pulse rates faster than those produced by their males, whereas in the intermediate and slowest pulse rate species there was no significant difference between the male trait and the female response. This work suggests that species with the most extreme signal in the community respond to a greater range of signals, potentially resulting in a higher probability of hybridization during secondary contact, and revealing interactions between mate recognition and other aspects of sexual selection.

Contributions of natural and sexual selection to the evolution of premating reproductive isolation: a research agenda.

Speciation by divergent natural selection is well supported. However, the role of sexual selection in speciation is less well understood due to disagreement about whether sexual selection is a mechanism of evolution separate from natural selection, as well as confusion about various models and tests of sexual selection. Here, we outline how sexual selection and natural selection are different mechanisms of evolutionary change, and suggest that this distinction is critical when analyzing the role of sexual selection in speciation. Furthermore, we clarify models of sexual selection with respect to their interaction with ecology and natural selection. In doing so, we outline a research agenda for testing hypotheses about the relative significance of divergent sexual and natural selection in the evolution of reproductive isolation.

Diversification under sexual selection: the relative roles of mate preference strength and the degree of divergence in mate preferences.

The contribution of sexual selection to diversification remains poorly understood after decades of research. This may be in part because studies have focused predominantly on the strength of sexual selection, which offers an incomplete view of selection regimes. By contrast, students of natural selection focus on environmental differences that help compare selection regimes across populations. To ask how this disparity in focus may affect the conclusions of evolutionary research, we relate the amount of diversification in mating displays to quantitative descriptions of the strength and the amount of divergence in mate preferences across a diverse set of case studies of mate choice. We find that display diversification is better explained by preference divergence rather than preference strength; the effect of the latter is more subtle, and is best revealed as an interaction with the former. Our findings cast the action of sexual selection (and selection in general) in a novel light: the strength of selection influences the rate of evolution, and how divergent selection is determines how much diversification can occur. Adopting this view will enhance tests of the relative role of natural and sexual selection in processes such as speciation.

Species recognition and patterns of population variation in the reproductive structures of a damselfly genus.

The selection pressures imposed by mate choice for species identity should impose strong stabilizing selection on traits that confer species identity to mates. Thus, we expect that such traits should show nonoverlapping distributions among closely related species, but show little to no variance among populations within a species. We tested these predictions by comparing levels of population differentiation in the sizes and shapes of male cerci (i.e., the clasper structures used for species identity during mating) of six Enallagma damselfly species. Cerci shapes were nonoverlapping among Enallagma species, and five of six Enallagma species showed no population variation across their entire species ranges. In contrast, cerci sizes overlapped among species and varied substantially among populations within species. These results, taken with previous studies, suggest that cerci shape is a primary feature used in species recognition used to discriminate conspecific from heterospecifics during mating.