Bat Calls: How to Identify Species by Sound

Identifying bat species by their echolocation calls is a practical skill for students, researchers, and wildlife professionals who conduct acoustic surveys. Bats produce ultrasonic calls that are inaudible to human ears, but bat detectors convert these signals into audible or visual forms that can be analyzed. Species identification from calls relies on measuring frequency, duration, and call shape, then comparing those measurements against reference libraries. This article explains how echolocation calls are structured, how to record and analyze them, and what limitations affect accurate identification.

What Bat Calls Are and Why They Matter

Bat echolocation calls are high-frequency sound pulses that bats emit to navigate and locate prey. These calls reflect off objects and return as echoes, which the bat interprets to build a spatial picture of its surroundings. The calls are species-specific to a degree, meaning that trained observers and computer programs can often identify which bat species produced a given call.

Acoustic monitoring has become a standard tool for studying bat populations. Researchers use bat detectors to record calls in the field, then analyze the recordings to determine which species are present in an area. This approach is less invasive than capturing bats and allows surveys to cover large areas over long periods. Acoustic data can reveal habitat use patterns, activity levels, and the presence of rare or threatened species.

The practical value of acoustic identification extends to conservation and land management. Knowing which bat species use a particular forest, wetland, or agricultural landscape helps managers make informed decisions about habitat protection, wind turbine placement, and forestry practices. For example, studies of Bechstein's bats in forest wind parks show that these bats prefer high-quality foraging habitat near their roosts but avoid turbines when farther away, information that can guide turbine siting and operation schedules (Habitat use of Bechstein's bats within wind parks in forests).

How Echolocation Calls Are Structured

Bat calls vary in frequency, duration, and shape. Understanding these structural elements is the foundation of acoustic identification.

Call Types

Bat calls fall into broad structural categories based on how frequency changes over time.

Frequency modulated calls sweep across a range of frequencies, usually downward. These calls provide detailed information about the environment because the bat can compare echoes at different frequencies. Most bat species in temperate regions emit frequency modulated calls.

Constant frequency calls maintain a single frequency for most of the call duration. Horseshoe bats and leaf-nosed bats use constant frequency calls, often combined with a short frequency modulated component at the start and end. The constant frequency portion is useful for detecting fluttering prey because the bat can detect Doppler shifts in the returning echo.

Quasi constant frequency calls are intermediate, with a shallow frequency sweep. Several vesper bat species produce these calls, which are often described as frequency modulated to quasi constant frequency.

A study of eight bat species in southern Hokkaido, Japan, categorized calls into three types: frequency modulated to constant frequency to frequency modulated for the greater horseshoe bat, frequency modulated for four species, and frequency modulated to quasi constant frequency for three species (Acoustic identification of eight species of bat inhabiting forests of southern Hokkaido, Japan). This categorization helped researchers distinguish species before applying statistical analysis.

Measurable Parameters

When analyzing a bat call, several parameters are measured from the sonogram, which is a visual display of frequency over time.

Peak frequency is the frequency with the most energy in the call. This is often the most useful parameter for distinguishing species because it varies predictably among species.

Start frequency and end frequency describe the range of the frequency sweep.

Call duration is the length of the call in milliseconds.

Inter-call interval is the time between successive calls.

Bandwidth is the difference between the highest and lowest frequencies in the call.

A study of two morphologically similar bat species, Miniopterus magnater and Miniopterus fuliginosus, found that although call durations were similar, start frequency, end frequency, and peak frequency differed significantly between the species (Acoustic identification of two morphologically similar bat species). Using spectral features alone, 92.3% of calls were attributed to the correct species.

At a Glance: Bat Call Characteristics by Species Group

The following table summarizes typical call characteristics for common bat groups. These values are general reference points, not definitive identification criteria. Actual calls vary by region, habitat, and individual bat.

The call characteristics in this table are drawn from the general patterns described in acoustic identification studies. For example, the study of hipposiderid bats in northern Vietnam found that six species in the family Hipposideridae were distinct in their acoustic characteristics, allowing field identification (New distributional records and acoustic identification of hipposiderid bats from Son La and Lang Son provinces, northern Vietnam).

Recording Bat Calls in the Field

Successful acoustic identification begins with good recordings. The equipment you use and the way you deploy it affect the quality of the data you collect.

Bat Detector Types

Bat detectors convert ultrasonic calls into signals that humans can perceive or that can be recorded for later analysis.

Heterodyne detectors tune to a single frequency and make calls near that frequency audible. These detectors are useful for real-time monitoring but capture only a narrow frequency band, so they miss calls outside the tuned range.

Frequency division detectors divide the frequency of incoming calls by a fixed factor, making them audible. These detectors preserve timing information but lose some frequency detail.

Time expansion detectors record a short segment of sound and play it back at a slower speed. This method preserves the full frequency content of the call, making it suitable for detailed analysis.

Full spectrum detectors digitize the entire ultrasonic signal directly. These detectors produce the highest quality recordings for computer analysis.

The choice of detector affects the accuracy of species identification. A study of automated classifiers for northeastern United States bat species found that one program performed less accurately than others, likely because it operated on zero-crossing data and was less accurate for recordings converted from full-spectrum to zero-crossing format (Automated echolocation classifiers vary in accuracy for northeastern U.S. bat species). Full-spectrum recordings generally provide more information for analysis.

Recording Protocol

Standardized recording protocols improve the comparability of data across sites and studies. The Sonozotz project in Mexico assembled a nation-wide library of bat echolocation calls using a standardized protocol that covered different recording habitats, recording techniques, and call variation inherent to individuals (The Sonozotz project: Assembling an echolocation call library for bats in a megadiverse country). The project included 69 species of echolocating bats, representing 50% of bat species found in the country.

When setting up a recording station, consider the following factors.

Placement affects detection range. Place detectors in open areas where bats are likely to fly, such as over water, along forest edges, or near roost entrances.

Height matters because some species forage at canopy level while others fly close to the ground.

Timing is critical. Most bat activity occurs after sunset and before sunrise, with peaks in the first few hours after dark.

Weather affects call transmission. Temperature and relative humidity influence atmospheric attenuation, which is the loss of sound energy as it travels through air. Atmospheric attenuation is strongly positively correlated with call frequency, meaning high-frequency calls lose energy faster than low-frequency calls (Weather conditions determine attenuation and speed of sound). Variable weather conditions result in variable and unknown effects on the recorded call, affecting estimates of call frequency and intensity, particularly for high frequencies.

Recording from Captured or Hand-Released Bats

For building reference libraries, researchers often record calls from bats that have been captured and identified to species. These recordings provide known-identity calls that can be used to train identification models.

A study in the Brazilian Amazon used recordings of free-flying and hand-released bats to train an automatic classification algorithm (Stronger together: Combining automated classifiers with manual post-validation optimizes the workload vs reliability trade-off of species identification in bat acoustic surveys). The study found that random forest models confirmed the reliability of using calls of both free-flying and hand-released bats to train custom-built automatic classifiers.

Analyzing Bat Calls for Species Identification

Once you have recorded calls, the next step is to analyze them. This involves measuring call parameters and comparing them against reference data.

Manual Analysis

Manual analysis involves viewing sonograms and measuring call parameters by eye or with software tools. This approach requires training and experience but allows the analyst to account for call quality and context.

A study of eight bat species in Hokkaido, Japan, used sonograms of 171 calls from eight species and categorized them into three types by structure (Acoustic identification of eight species of bat inhabiting forests of southern Hokkaido, Japan). The calls of the greater horseshoe bat could easily be distinguished from all other species by eye. For the remaining calls, seven parameters were examined using discriminant function analysis, and 92% of calls were correctly classified to species.

Automated Classification

Automated classifiers use computer algorithms to identify species from recorded calls. These programs are widely used because they can process large datasets quickly.

A study of automated classifiers for northeastern United States bat species tested three programs: Bat Call Identification, Kaleidoscope Pro, and SonoBat (Automated echolocation classifiers vary in accuracy for northeastern U.S. bat species). The study used 1,500 full-spectrum reference calls with known identities for nine species. The results showed that negative predictive value and specificity were high across all species categories for SonoBat and Kaleidoscope Pro, indicating success at avoiding false positives. However, positive predictive value and sensitivity were relatively low, particularly for individual Myotis species, indicating these programs are prone to false negatives.

The study cautioned that a bat acoustic expert should verify automatically classified files when making species-specific regulatory or conservation decisions.

Statistical Classification Methods

Researchers have used various statistical methods to classify bat calls to species. Discriminant function analysis and artificial neural networks are two common approaches.

A study of 14 sympatric bat species in Britain recorded echolocation calls and measured one temporal and four spectral features from each call (Acoustic identification of twelve species of echolocating bat by discriminant function analysis and artificial neural networks). Discriminant function analysis achieved an overall correct classification rate of 79%, while an artificial neural network achieved 87%. This was the first published study to use artificial neural networks to classify the echolocation calls of bats to species level.

The Role of Call Libraries

Call libraries are collections of recorded calls with known species identities. These libraries serve as reference data for training and testing identification models.

The ChiroVox website was created to facilitate the sharing of bat sound recordings together with their metadata, including biodiversity data and recording circumstances (ChiroVox: a public library of bat calls). To date, more than 30 researchers have contributed over 3,900 recordings of nearly 200 species, making ChiroVox the largest open-access bat call library currently available. Each recording has a unique identifier that can be cited in publications, making acoustic analyses repeatable.

The Sonozotz project represents the first nation-wide library of bat echolocation calls for a megadiverse country (The Sonozotz project: Assembling an echolocation call library for bats in a megadiverse country). The project included recommendations on how the database can be used and how the sampling methods can be replicated in countries with similar environmental and geographic conditions.

Practical Workflow for Acoustic Surveys

A structured workflow helps ensure that acoustic surveys produce reliable data. The following steps outline a practical approach.

Step 1: Define Survey Objectives

Determine what questions the survey will answer. Are you documenting species presence in an area? Comparing bat activity across habitat types? Monitoring a rare species? The objectives determine the sampling design, equipment choices, and analysis methods.

Step 2: Select Equipment and Recording Settings

Choose detectors appropriate for the target species and analysis method. Full-spectrum detectors are recommended when detailed frequency analysis is needed. Set recording schedules to cover the expected activity period, typically from sunset to sunrise.

Step 3: Deploy Detectors

Place detectors at representative locations within the study area. Record the location, date, time, weather conditions, and habitat characteristics for each deployment. Consistent metadata allows for meaningful comparisons across recordings.

Step 4: Collect and Organize Recordings

Transfer recordings from detectors to a computer and organize them by site, date, and deployment. Use a consistent file naming convention that includes location and date information.

Step 5: Analyze Recordings

Process recordings using manual or automated methods. For automated classification, run the recordings through the chosen software and review the output. For manual analysis, view sonograms and measure call parameters.

Step 6: Validate Results

Automated classifications should be verified, especially when the results will inform management decisions. A study of automated classifiers found that a bat acoustic expert should verify automatically classified files when making species-specific regulatory or conservation decisions (Automated echolocation classifiers vary in accuracy for northeastern U.S. bat species).

Step 7: Document and Report

Record the methods, results, and limitations of the survey. Include information about the equipment, recording settings, analysis methods, and any quality control measures applied.

Records and Measurements for Acoustic Surveys

Maintaining detailed records is essential for acoustic surveys. The following measurements and observations should be documented for each recording session.

Site information includes location coordinates, habitat type, and proximity to water, roosts, or other features that influence bat activity.

Weather conditions affect call transmission and detection. Temperature and relative humidity influence atmospheric attenuation, which limits the distance at which calls can be detected (Weather conditions determine attenuation and speed of sound). Wind speed and precipitation also affect recording quality.

Equipment settings include detector type, gain, trigger sensitivity, and recording schedule. Consistent settings across deployments improve comparability.

Call quality should be assessed for each recording. Calls that are faint, partially recorded, or obscured by noise may not be suitable for species identification.

Species identifications should be recorded with confidence levels. Distinguish between calls that are confidently identified, calls that are identified to genus or species group, and calls that cannot be identified.

A study of automated classifiers for rare bat species found that the total number of audio files recorded within a night had a great effect on whether a rare species was detected (Maximum likelihood estimators are ineffective for acoustic detection of rare bat species). Both programs performed poorly with determining presence for any species at low species ratio, below 25%. This finding highlights the importance of recording effort and species composition when interpreting acoustic survey results.

Common Failure Patterns in Acoustic Identification

Several recurring problems can compromise the accuracy of acoustic identification. Recognizing these failure patterns helps researchers avoid them.

Overreliance on Automated Classifiers

Automated classifiers are convenient but not infallible. A study of northeastern United States bat species found that positive predictive value and sensitivity were relatively low for automated programs, particularly for individual Myotis species (Automated echolocation classifiers vary in accuracy for northeastern U.S. bat species). The programs were prone to false negatives, meaning they missed calls that were actually present.

Inadequate Reference Libraries

Identification models are only as good as the reference data used to train them. A study of call libraries and acoustic filters found that overall correct classification rates were significantly lower, 15% to 23% lower, when tested on a second call library instead of the training library (The effect of call libraries and acoustic filters on the identification of bat echolocation). This finding indicates that it may not be possible to extend inferences from one library to another without validation.

Geographic Variation in Calls

Bat calls vary geographically, even within the same species. A study of echolocation plasticity in Pipistrellus species found that bats in allopatry, meaning populations that do not share a range, showed call differences that complicate acoustic identification (Bat echolocation plasticity in allopatry: a call for caution in acoustic identification of Pipistrellus sp.). Reference libraries from one region may not accurately represent calls from another region.

Social Calls and Contextual Variation

Bats adjust their echolocation calls in different social contexts. A study of intermediate leaf-nosed bats found that a single bat individual altered its peak frequency of echolocation calls to approach the call of new colony members, and two bats from different colonies adjusted their call frequencies toward each other to a similar frequency after being chronically cohabitated (Compromise in echolocation calls between different colonies of the intermediate leaf-nosed bat). These results indicate that compromise in echolocation calls might be used to ensure effective mutual communication among bats.

Weather Effects on Call Transmission

Weather conditions affect the transmission of bat calls through the atmosphere. Atmospheric attenuation is a nonlinear function of call frequency, temperature, relative humidity, and atmospheric pressure (Weather conditions determine attenuation and speed of sound). Variable weather conditions result in variable and unknown effects on the recorded call, affecting estimates of call frequency and intensity, particularly for high frequencies.

Limitations of Acoustic Identification

Acoustic identification has inherent limitations that researchers must acknowledge when interpreting results.

Species with Similar Calls

Some bat species produce calls that are difficult to distinguish acoustically. Morphologically similar species may have overlapping call characteristics. A study of two Miniopterus species found that although call durations were similar, there were significant differences in start, end, and peak frequencies between the species (Acoustic identification of two morphologically similar bat species). However, not all similar species can be distinguished this reliably.

Individual and Contextual Variation

Individual bats vary in their call characteristics, and the same bat may produce different calls in different contexts. Calls recorded from bats flying in confined spaces differ from calls recorded from bats flying in open areas. This variation can complicate species identification.

Detection Range and Sampling Bias

Bat detectors have limited detection ranges, and these ranges vary by call frequency and weather conditions. High-frequency calls attenuate more rapidly than low-frequency calls, meaning that species with high-frequency calls are detected at shorter distances. This creates a sampling bias that can affect species composition estimates.

The Need for Expert Verification

Given the limitations of automated classification, expert verification is often necessary. A study of automated classifiers concluded that a bat acoustic expert should verify automatically classified files when making species-specific regulatory or conservation decisions (Automated echolocation classifiers vary in accuracy for northeastern U.S. bat species). This recommendation reflects the reality that automated programs, while useful, cannot fully replace human expertise.

Welfare and Safety Considerations

Acoustic monitoring is a non-invasive method that does not require capturing or handling bats. This is an advantage over methods that involve physical capture, which can cause stress to bats and requires permits and training.

However, researchers should be aware of the welfare implications of their work. When recording calls from hand-released bats, researchers must follow ethical guidelines for handling wild animals. Captured bats should be processed quickly and released at the capture site.

Safety considerations apply to fieldwork conducted at night. Researchers should work in teams, carry appropriate lighting, and be aware of their surroundings. Sites may include caves, mines, or other hazardous locations.

Professional Escalation Criteria

Knowing when to seek expert assistance is important for maintaining data quality. The following situations warrant consultation with a bat acoustic specialist.

Unusual or ambiguous calls that do not match reference libraries should be reviewed by an expert before being assigned to a species.

Species of conservation concern require verification of acoustic identifications. Misidentification of a threatened species can have regulatory consequences.

Regulatory or legal decisions based on acoustic data should include expert verification. Federal guidance in the United States requires the use of automated identification programs for determining presence or probable absence of threatened and endangered bats (Maximum likelihood estimators are ineffective for acoustic detection of rare bat species).

Cross-regional comparisons require careful validation. Call libraries from one region may not accurately represent calls from another region, and expert input may be needed to interpret cross-regional data.

New or understudied species may not be represented in existing reference libraries. Expert assistance may be needed to confirm identifications or to add new recordings to reference collections.

Frequently Asked Questions

What equipment do I need to record bat calls?

You need a bat detector that can convert ultrasonic calls into recordable signals. Full-spectrum detectors produce the highest quality recordings for species identification. You also need a recording device and software for viewing and analyzing sonograms. The choice of detector affects the accuracy of species identification, so select equipment appropriate for your research questions.

How do I measure bat call frequency?

Bat call frequency is measured from a sonogram, which displays frequency on the vertical axis and time on the horizontal axis. The peak frequency is the frequency with the most energy in the call. Start frequency and end frequency describe the range of the frequency sweep. These measurements are typically made using acoustic analysis software.

Can I identify bats by ear?

No. Bat echolocation calls are ultrasonic, meaning they are above the range of human hearing. Bat detectors convert these calls into audible signals, but the converted sounds do not preserve the original frequency information. Species identification requires analysis of the recorded calls using sonograms or automated classification software.

How accurate are automated bat call classifiers?

Accuracy varies by program, species, and recording conditions. A study of northeastern United States bat species found that automated programs had high negative predictive value and specificity but relatively low positive predictive value and sensitivity, particularly for individual Myotis species (Automated echolocation classifiers vary in accuracy for northeastern U.S. bat species). Expert verification is recommended for species-specific decisions.

What is a call library and why is it important?

A call library is a collection of recorded bat calls with known species identities. These libraries serve as reference data for training and testing identification models. The ChiroVox website is the largest open-access bat call library currently available, with over 3,900 recordings of nearly 200 species (ChiroVox: a public library of bat calls). Call libraries are essential for acoustic identification because they provide the reference data needed to distinguish species.

How does weather affect bat call recordings?

Weather conditions affect the transmission of sound through the atmosphere. Atmospheric attenuation is strongly positively correlated with call frequency, and it is also significantly influenced by temperature and relative humidity (Weather conditions determine attenuation and speed of sound). Variable weather conditions result in variable and unknown effects on the recorded call, affecting estimates of call frequency and intensity, particularly for high frequencies.

Can bat calls vary within a species?

Yes. Bat calls vary geographically, individually, and contextually. A study of intermediate leaf-nosed bats found that bats adjusted their call frequencies when encountering bats from different colonies (Compromise in echolocation calls between different colonies of the intermediate leaf-nosed bat). Geographic variation in calls means that reference libraries from one region may not accurately represent calls from another region.

When should I consult a bat acoustic expert?

Consult an expert when you encounter unusual or ambiguous calls, when working with species of conservation concern, when making regulatory or legal decisions based on acoustic data, or when conducting cross-regional comparisons. Automated classifiers should be verified by an expert when making species-specific regulatory or conservation decisions (Automated echolocation classifiers vary in accuracy for northeastern U.S. bat species).

Related Articles

• Rna Codon Chart
• Rna Codon Chart
• Rna Codon Chart
• plant disease identification
• Nature Cell Biology

References and Further Reading

• NCBI Literature Resources. National Center for Biotechnology Information.
• PubMed. National Library of Medicine.
• The effect of call libraries and acoustic filters on the identification of bat echolocation.. Ecology and evolution, 2014.
• ChiroVox: a public library of bat calls.. PeerJ, 2022.
• The Sonozotz project: Assembling an echolocation call library for bats in a megadiverse country.. Ecology and evolution, 2020.
• Automated echolocation classifiers vary in accuracy for northeastern U.S. bat species.. PloS one, 2024.
• 'Compromise' in Echolocation Calls between Different Colonies of the Intermediate Leaf-Nosed Bat (Hipposideros larvatus).. PloS one, 2016.
• Acoustic identification of twelve species of echolocating bat by discriminant function analysis and artificial neural networks.. The Journal of experimental biology, 2000.
• Weather conditions determine attenuation and speed of sound: Environmental limitations for monitoring and analyzing bat echolocation.. Ecology and evolution, 2018.
• Acoustic identification of eight species of bat (mammalia: chiroptera) inhabiting forests of southern hokkaido, Japan: potential for conservation monitoring.. Zoological science, 2004.
• Bayesian network 3D event reconstruction in the Cygno optical TPC for dark matter direct detection.. 2025.
• Surface geoenvironments shape organic matter inputs in iron-ore caves of Carajás, southeastern Amazonia.. 2026.
• MQADet: a plug-and-play paradigm for enhancing open-vocabulary object detection via multimodal question answering.. 2026.
• Habitat use of Bechstein´s bats (Myotis bechsteinii) within wind parks in forests.. 2026.
• Anti-Atherogenic Actions of Pomegranate Polyphenol Punicalagin and Its Metabolites: In Vitro Effects on Vascular Cells and In Vivo Atheroprotection by Urolithin A via Anti-Inflammatory and Plaque-Stabilising Mechanisms.. 2026.
• The effect of Gougunao tea polysaccharide on lipid metabolism in hyperlipidemia induced by a high-fat diet and its structural characteristics.. 2025.
• Development of a handheld chlorophyll content detector on wheat and maize leaves based on RGB sensor.. 2025.
• Acoustic identification of two morphologically similar bat species, Miniopterus magnater and Miniopterus fuliginosus (Chiroptera, Miniopteridae). 2019.
• Maximum likelihood estimators are ineffective for acoustic detection of rare bat species. PLoS ONE, 2025.
• Automated acoustic identification of nine bat species of the eastern United States. 2007.
• Stronger together: Combining automated classifiers with manual post-validation optimizes the workload vs reliability trade-off of species identification in bat acoustic surveys. Ecological Informatics, 2019.
• NEW DISTRIBUTIONAL RECORDS AND ACOUSTIC IDENTIFICATION OF HIPPOSIDERID BATS (CHIROPTERA: HIPPOSIDERIDAE) FROM SON LA AND LANG SON PROVINCES, NORTHERN VIETNAM. Journal of Science Natural Science, 2025.
• Correction to: Bat echolocation in continental China: a systematic review and first acoustic identification key for the country. Mammal Research, 2021.
• Variability in echolocation call design of 26 Swiss bat species: Consequences, limits and options for automated field identification with a synergetic pattern recognition approach. Mammalia, 2004.
• Bat echolocation plasticity in allopatry: a call for caution in acoustic identification of Pipistrellus sp.. Behavioral Ecology and Sociobiology, 2021.
• Testing the performances of automated identification of bat echolocation calls: A request for prudence. Ecological Indicators, 2017.

This article is educational and does not replace institutional policy, professional advice, or applicable safety and regulatory requirements.