Skip to main content
NCBI home page
PLOS One logoLink to PLOS One
. 2026 Oct 26;21(10):e0360298. doi: 10.1371/journal.pone.0360298

Home security cameras as a tool for behavior observations and science affordability

Billie C Goolsby 1,☯, Marie-Therese Fischer 1,☯, Tony G Chen 2, Daniela Pareja-Mejía 1,3, Daniel A Shaykevich 1, Amaris R Lewis 1, Gaelle Raboisson 1, Katharina Dellefont 1, Madison P Lacey 1, Lauren A O'Connell 1,*
Editor: Patrick R Stephens4
PMCID: PMC13704821  PMID: 42932364

Abstract

Reliably capturing transient animal behavior in the field and laboratory remains a logistical and financial challenge, especially for small ectotherms. Here, we evaluate home security cameras as affordable, accessible and suitable alternatives for monitoring small, cold-blooded animals historically overlooked by commercial camera traps. These cameras are often lightweight, weather-resistant, operate offline or online, store data continuously for up to four weeks and can use phone notifications over local Wi-Fi to alert users to behaviorally relevant events. We first benchmark a representative home security camera against an established trail camera under matched laboratory conditions in both continuous and motion-triggered modes, comparing event detection, recording duration, and missed-event rates. We then validate the system during a 1.5-month deployment in a tropical primary rainforest, assessing weather durability, sensitivity-versus-data-volume trade-offs, and whether behavioral measurements obtained in the laboratory are reproducible under natural conditions. We discuss the relative affordability of our system for researchers in the Global South, home to the largest ectotherm diversity.

Introduction

Accurately quantifying animal behavior is critical for understanding how organisms respond to challenges and opportunities in their environment, which influences the fitness and survival of a species. Human presence can obscure natural behaviors [1, 2], and remote cameras bypass the need for physical presence of the researcher and allows for operationalization of behaviors [3]. Remote camera approaches are used to study behavior at both population [4, 5] and individual levels [6, 7]. Mammals and birds are the most common fauna monitored with cameras in the wild, but only a small range of prior studies (<2%) have focused on ectotherms, specifically amphibians and reptiles [8–10]. This gap between ectotherms and endotherms can be explained by both technical and logistical difficulties: many camera traps rely on surface body heat signatures (passive infrared, PIR) warmer than the surrounding environment [11]. Ectotherms do not produce endogenous body heat higher than the surrounding environment, making PIR camera traps unreliable without experimental background modifications for maximizing heat contrast [10, 12, 13], which is difficult in remote areas or when minimal environmental interference is desired. Moreover, countries in the Global South hold much of the world’s ectotherm biodiversity [14], yet accessibility and affordability of camera traps for local researchers are prohibitive for this research tool [15]. Indeed, tropical biomes, which are at highest defaunation risk, are disproportionately less monitored with camera traps compared to other environments [15]. Increasing camera weather resistance correlates with higher prices, and many camera traps rely on cellular networks headquartered in the Global North, which can present difficulties in network-limited landscapes [16]. An ideal camera system should be affordable, capable of detecting small ectotherm movement, weather-resistant, and operational in the absence of electricity and phone connectivity.

Home security cameras, originally designed for the residential market, offer an affordable alternative class of recording device that can be repurposed for research. These cameras frequently use pixel change detection, which compares pixel values between successive video frames, rather than infrared detection, to trigger recordings [17]. In daylight, home security cameras use natural lighting for pixel comparisons, and at night, use an active IR system converted into a greyscale image used for pixel change. This enables detection of ectotherm movement, whose body temperature does not differ markedly from ambient conditions. Several home security models are weather-resistant, operate offline, and connect to local Wi-Fi networks rather than depending on cellular providers, which is a key advantage in tropical field sites where Global North cellular coverage is poor or absent.

Establishing whether such cameras are suitable for ectotherm behavioral research requires two complementary lines of evidence. First, the affordable camera must perform comparably to established commercial systems under controlled conditions, where independently verified behavior is available for direct comparison. Second, the camera must remain functional and produce ecologically valid data under the real-world conditions, including humidity and temperature extremes, offline operation, and free-ranging animals in complex natural backgrounds.

Here, we provide a technical comparison of seven camera systems currently available on the United States market, including affordable home security cameras and camera traps commonly used for research. We then present a two-stage evaluation of a representative system (Wyze v3) for ectotherm behavioral research. First, we benchmark the security camera against an established commercial trail camera (GardePro A3S) in a laboratory terrarium to determine (i) how the two systems compare in the number of behavioral events detected, (ii) how the cameras compare in the duration of behavior recorded per event, and (iii) what fraction of events are recorded by one camera and missed by the other (Study 1). We next validated the system in the field during a 1.5-month deployment in a tropical primary rainforest in a rubric with three independent criteria: hardware survival under tropical conditions, the trade-off between motion-detection sensitivity and event capture, and the ecological validity of laboratory-measured behavioral metrics by comparing paternal visit duration in poison frogs between captive and free-ranging individuals (Study 2).

We used two poison frog species as study systems to address different aspects of camera-assisted behavioral monitoring: the relatively large dyeing poison frog (Dendrobates tinctorius) served as our laboratory benchmarking system because its larger vivaria could accommodate both camera systems simultaneously, allowing direct comparison while monitoring adult behavior. The variable poison frog (Ranitomeya variabilis) provided a field-relevant system for monitoring parental care and tadpole development. Males transport newly hatched tadpoles to small water-filled leaf axils, where the cannibalistic tadpoles develop individually. These discrete, abundant, and accessible aquatic nurseries provide natural observational units that can be readily monitored and manipulated, making R. variabilis particularly suitable for evaluating camera-based monitoring of parental care and tadpole behavior under field conditions.

We hypothesized that the Wyze v3 would outperform the GardePro A3S in detecting and recording ectotherm behavioral events because its pixel-change detection is better suited to small ectotherms than passive infrared triggering. However, given its substantially lower cost and design as a consumer home security camera rather than a research-grade camera trap, we expected this improved behavioral performance to come at the cost of reduced hardware reliability under tropical field conditions. We further predicted that behavioral measurements obtained with pixel-change detection in captivity would be comparable to those obtained from free-ranging animals.

Materials and methods

Study 1: Camera comparison and laboratory benchmarking

Camera systems compared.

We compiled technical specifications and availability of seven camera systems (Table 1) to evaluate their suitability for lab and field studies of amphibian behavior: three commercial trail or monitoring cameras (Reconyx Hyperfire 2 [Holmen, WI, USA]; Bushnell Nature View HD [Overland Park, KS, USA], used in [10]; GardePro A3S [Hong Kong], used in [18]), three home security cameras (Wyze v3 [Seattle, WA, USA], Google Nest [Mountain View, CA, USA], used in [19, 20]; Tapo C120 [Irvine, CA, USA]), and a manual handheld camera (Sony HDR-CX405 [Tokyo, Japan], used in [21, 22] respectively. To select two cameras for direct benchmarking, our criteria were ectotherm detectability (pixel-change vs. passive infrared sensors (PIR)), affordability, and local-network availability. We opted to compare the Wyze v3 and GardePro A3S camera systems, as they were the most affordable representatives of their respective classes at the time of the study, while sharing comparable motion-detection length, storage capabilities, and weather resistance.

Table 1. Camera performance for studying ectotherm behavior.
System Specifications Consequence on Amphibian Studies Wyze Camera v3* Google Nest Camera, Indoor & Wired Tapo C120 Sony Camera HDR-CX405

 Reconyx Hyperfire 2 Cellular Professional Covert IR Camera Bushnell Nature View HD GardePro A3S Trail Camera
Ability to notify phone with local network Quickly alerts researchers to behavior Yes Yes Yes No Yes No No
Ability to notify phone without local network Quickly alerts researchers to behavior No No No No No No No
Cellular Source Alerts rely on either local networks or cellular ones 2.4 GHz local Wifi 2.4GHz/5GHz Wi-Fi IEEE 802.11b/g/n, 2.4 GHz Wi-Fi None 3-4G LTE (ATT & Verizon, some international plans) None None
Price (USD) Cost influences number of cameras purchasable $39.98 $99.99 $35.99 $229.99 $459.99 $299.99 $89.99
Image Resolution Equivalent across most cameras HD 1920 x 1080p HD (1920 x 1080) 2K QHD 4MP (2560 × 1440 px) 2.1 Megapixels (1920 x 1080) 1080P HD Widescreen HD 1080P 48 Megapixels

HD
Video Quality and sound Equivalent across most cameras HD 1920 x 1080p HD (1920 x 1080) 2K QHD 4MP (2560 × 1440 px, 20 fps) 1920 x 1080/60p 720P HD Video with Audio (cellular enabled cameras do not transmit video) HD 1920x1080p 2304 x 1296P@20fps (16:9

OR

1920 x 1080P@30fps (16:9)

OR

1280 x 720P@30fps (16:9)
Detection Method (Day) Risk of not detecting amphibians with PIR Pixel change Pixel change 1/2.9” Progressive Scan CMOS Starlight Sensor Exmor R CMOS Sensor PIR PIR PIR (high, normal low)
Motion Detection Video length Variable, depends on research goal Up to 5 minutes per notification Up to 3 hours of event history totally Up to 3 minutes per notification None, manual Up to 90 seconds per notification up to 60s Up to 5 minutes
Continuous Video length Remote monitoring of behaviors that may last longer than several minutes. 24/7 for up to 4 weeks with SD card 24/7 for up to 10 days with subscription 24/7 for up to 4 weeks with SD card Approximately 2 hrs 35 minutes of battery life No continuous recording abilities

(trigger only)		 No continuous recording abilities (trigger only) No continuous recording abilities

(trigger only)
Detection Method (Night and wavelength in nm) IR wavelength that can detect amphibians Near-IR (940) and Far-IR (850) Far-IR (850) Near-IR (940) and Far-IR (850) None explicitly stated No-Glow™ High Output Covert IR PIR PIR (940nm) No Glow IR LEDs
Camera weight (g) Lower weights are more carryable for fieldwork 98.8 393 100 190-215 ~750 246.6 530
IP Score IP66 best for heavy pressurized water exposure, IP65 handles rain IP65 None provided. IP66 None provided. IP66 IP66 IP66
Operating Temperatures (°C) Most cameras tolerate most biome temperature ranges −20 - + 40 0- + 40 −25 - + 45 None provided. −40 - + 60 −4 - + 140 −20 - + 60
Night vision distance (m) Depends on animal of interest None provided 15 30 None provided. 45 18.3 30
Cloud saving abilities Depends on researcher needs Yes, with extra fees Yes, with extra fees Yes, with extra fees No No No No
Highest amount of internal memory allowed on camera via SD card (Gb) Higher Gb allows for cameras being left alone longer 512 Up to 1 hr of internal storage (no Gb amount provided). 512 No limit for SD card 512 32 512
Power source (years last) USB requires continuous power source, which needs to be resolved before field work USB USB USB Battery or USB 12 AA (2 yrs)

AA (4–12)

Up to one year		 AA (8)

Over a year depending on usage
Field of View Depends on camera’s ability to orient and point towards area of interest 130° 135° 120°(Diagonal), 103°(Horizontal), 55°(Vertical) 90-270° None provided 50° Total 120° Central zone: 60°, Left side: 30°,Right side: 30°
Open in a new tab

We compare features and accessibility of three home security systems (Wyze, Google Nest, Tapo) against four traditional camera traps for animal behavior currently on the United States commercial market (Sony, Reconyx Hyperfire 2, Bushnell Nature View HD, GardePro). If cameras are no longer commercially available by the original vendor, we provide the most similar commercially available camera from the same vendor. We discuss each feature in the context of studies on amphibian behavior. Abbreviations: LTE: long term evolution (a standard for wireless data transmission), HD: high definition, PIR: passive infrared sensor, IR: infrared, CMOS: Active pixel sensor; IP65 and IP66: indices for ingress protection (weather resistance, the higher the better), AA: single cell dry battery. Dark grey indicates home security cameras. Prices retrieved March 2026.

* discontinued and replaced with Wyze v4 as of March 2026.

Camera setup and modifications in a laboratory setting.

Detailed setup instructions are available on GitHub (https://github.com/laurenoconnelllab/WyzeSetUps) and a modification protocol for close-range lens refocus on protocols.io (dx.doi.org/10.17504/protocols.io.dm6gp3p38vzp/v1). In brief, Wyze v3 cameras were attached by velcro to the side of Exoterra tanks in the laboratory, around 15.5 to 17.5 cm above the focal area, powered via USB through a Nexwell charging station, and recorded to 128 GB secure digital (SD) cards (filled in approximately 10 days, then swapped) (Fig 2A).

Fig 2. Camera installations in lab vivaria and in the field.

Fig 2

Open in a new tab

(A) In lab vivaria, velcro was used to fix cameras above the desired focal subject, allowing for easy removal of the camera to switch SD cards. For monitoring parent-offspring interactions, the camera body (inlay, white arrow) was adjusted to focus on the tadpole nursery (black arrow). The USB cables were connected to an outlet via a USB charging station. Inlay pictures represent screenshots extracted from the recordings. (B) In their natural habitat, tadpoles grow up in water bodies formed by tightly overlapping leaf bases (black arrow, inlay) of large terrestrial plants called bromeliads. For observations, rubber bands were used to mount the cameras on bromeliad leaves above natural tadpole nurseries. The cameras (white arrows) were connected to the USB port of a power bank protected from the rain by two layers of plastic zip lock bags. (C) Visit length did not differ significantly between captive males or wild caregivers (purple; χ² = 0.42, p = 0.81, Kruskal-Wallis). (D) Cameras can be manually modified to improve resolution of focal sites, such as the tadpole inside of the plastic canister.

Anuran study system in the laboratory.

Dyeing poison frog (Dendrobates tinctorius). Laboratory benchmarking was performed with D. tinctorius as their vivaria were large enough to fit both camera types simultaneously (Fig 1). One captive-bred adult male and one adult female were housed together in a 45.72 × 45.72 × 45.72 cm Exoterra terrarium with sphagnum moss, driftwood, live Pothos plants, and water-filled petri dishes (R/O Rx, Josh’s Frogs). Frogs were fed Drosophila hydeii flies dusted with vitamin mix and Folsomia candida springtails three times per week. Both cameras simultaneously filmed a communal water pool and Coco-hut shelter. The observation housing was set on a 12:12 light cycle, the average temperature and humidity of the observation was recorded for each day of observation, usually around 23.5°C and 95% humidity within the tank. Recording comparisons between camera systems occurred in the same tank to standardize light, temperature, and humidity.

Fig 1. Comparisons of a home security camera (Wyze) in both continuous (24/7) and trigger formats versus a typical camera trap (GardePro) when recording the same area.

Fig 1

Open in a new tab

(A) Size comparison of GardePro and Wyze v3 to two species of poison frogs, Ranitomeya imitator (bottom) and Dendrobates tinctorius (top). Scale bar coincides to ~52 mm. (B) We recorded D. tinctorius around a communal water pool and Coco-hut for a fairer comparison between both camera systems. (C) Number of observations detected by Wyze in continuous mode and GardePro (no significant difference, ns). (D) Observation length in seconds per recording detected by Wyze in continuous mode and GardePro (no significant difference, ns). (E) Due to Wyze recording in continuous mode, no data can be lost due to errors in triggering. However, 24/7 collection of data can produce an abundance of false negative data that takes a long time to parse through. On average, GardePro loses around 20 seconds of animal behavior per trigger compared to 24/7 recordings. We then compared this to Wyze’s continuous mode, which is constantly saving data. (F) In continuous mode, Wyze collects a disproportionately large amount of false positive data where no behavior is actually recorded (333900 seconds) compared to actual behavior (6153 seconds). (G) Wyze was triggered less often than GardePro, however, (H) GardePro fails to detect the same event that Wyze records slightly more often than Wyze fails to detect the same event that GardePro records. (I) Of the same events that GardePro and Wyze both successfully detected, Wyze recorded a significantly longer interval of behavior (p < 0.001, V = 631, Wilcoxon signed rank test with continuity correction). To ensure equity in camera comparisons, we summed the amount of time GardePro or Wyze recorded total behavior per event, rather than per individual file, which would be due to storing mechanisms specific to each camera.

Laboratory benchmarking.

We benchmarked the Wyze v3 against the GardePro A3S in a laboratory terrarium housing a captive-bred adult male and adult female Dyeing poison frog (Dendrobates tinctorius). Both cameras simultaneously filmed the same focal area: a communal water pool and shelter. We ran two separate 96-h benchmarking periods: (i) Wyze in continuous (24/7) mode vs. GardePro in trigger-only mode (GardePro has no continuous option), and (ii) Wyze in trigger-only mode (75% motion sensitivity) vs. GardePro in trigger-only mode. For the continuous comparison, we quantified the number of behavioral observations per camera, the duration of behavior per observation, and the seconds of behavior lost per recording by GardePro relative to the continuously recording Wyze. We also measured the proportion of total Wyze recording time that contained no behavior. For the trigger-against-trigger comparison, we quantified the number of triggers per camera, the event-by-event concordance of detection (i.e., for each behavioral event, whether each camera detected it), and the duration of behavior recorded per event for events successfully detected by both cameras. To equitably compare cameras, we summed total recorded behavior per event, rather than per individual file, to control for camera-specific file-storage mechanisms.

Study 2: Field validation.

Laboratory benchmarking establishes performance under controlled conditions but does not establish whether a camera survives field deployment, whether the settings required to keep datasets manageable cause unacceptable event loss, or whether laboratory-derived behavioral measurements remain valid for free-ranging animals. The primary rainforest site is characterized by Clusia tree patches on bare granite rock with daily rainfall fluctuating between 0 and 30 mm and temperatures of 20°C to 49°C, measured at exposed rock surfaces in direct sunlight [23] (Fig. S3 in S1 File). The site is inhabited by a population of the variable poison frog Ranitomeya variabilis (CITES Appendix II, IUCN Least Concern), in which males defend territories and shuttle hatched tadpoles to bromeliad leaf-axil pools where the tadpoles develop. Male R. variabilis exhibit uniparental care (clutch attendance, tadpole transport, nursery surveillance) [24, 25]. This species was selected for field validation because of its small size (~18 mm) and because its parental behavior involves transient, hard-to-observe events occurring at small spatial scales, posing a technically and ecologically relevant challenge for camera systems.

Validation was structured around three independent criteria: (1) Hardware survival: we deployed 20 Wyze v3 units for 1.5 months and recorded the number of unit failures and their causes; (2) Sensitivity-versus-event-capture trade-off: we positioned ten cameras above bromeliad leaf-axils after temporarily removing inhabiting tadpoles to free preferred deposition sites, and independently verified deposition events by scoring the daily presence or absence of newly deposited tadpoles, then compared the count of independently verified events to the count captured at lowest motion sensitivity; (3) Ecological validity of laboratory-measured behavior: we compared male parental visit durations to tadpole nurseries between captive (laboratory colony) and wild (Inselberg) R. variabilis. Captive observations were obtained from two breeding pairs in 45.72 × 30.48 × 30.48 cm Exoterra (Rolf Hagen, Baie D’Urfé QC, Canada) terraria. Wild observations were obtained from monitoring five tadpoles in bromeliad leaf-axil pools for 10 days. Approval was granted by the scientific committee of the Nouragues Ecological Research Station and Stanford University IACUC (protocol #33691); captive recordings were approved under protocols #34242 and #34304.

Anuran study system in the field.

Variable poison frog (Ranitomeya variabilis). The comparison of laboratory and field studies was performed using R. variabilis. Adult males defend territories and shuttle hatched tadpoles to bromeliad pools where the cannibalistic tadpoles develop over around three months [24–26]. The species is CITES Appendix II, IUCN Least Concern. Captive frogs were housed in Exoterra terraria with film canisters as egg-deposition sites and water-filled canisters as tadpole nurseries. Wild frogs were monitored on the Inselberg of the Nouragues Nature Reserve, French Guiana (4°5’ N, 52°41’ W), a primary rainforest site.

Camera setup and testing of hardware survival under tropical conditions.

In the field, cameras were configured at the field station with Wi-Fi (motion-triggered mode, lowest sensitivity, sound detection and notifications off), then deployed offline. Cameras were secured to bromeliad leaves 20–40 cm above the water reservoir housing a tadpole using rubber bands. Units deployed on exposed rock surfaces were additionally shielded from rain with simple plastic strips fixed above the openings without covering them to prevent moisture ingress while not preventing rain from entering the tadpole nurseries. Cameras were powered by 26,800 mAh power banks wrapped in absorbent diapers sheltered by two layers of inexpensive plastic zip-lock bags with openings facing into opposite directions; this low-cost solution proved sufficient over the deployment period and could be replaced by outdoor waterproof boxes if needed (Fig 2B). SD cards and power banks were replaced every two days. A phone displaying the accurate time at startup was filmed to correct for the timestamp reset that occurred at each power cycle. Cameras recorded freely behaving animals in their home enclosures and in natural habitats and no animals were sacrificed or anesthetized in this study.

Sensitivity-versus-event-capture trade-off.

To quantify the cost of operating at lowest motion sensitivity (which is required to keep datasets manageable, since higher sensitivities trigger tens of thousands of recordings per day from light changes, moving vegetation, and rainfall), we positioned ten cameras above bromeliad leaf-axils after temporarily removing inhabiting tadpoles.

Ecological validity of laboratory-measured behavior.

We compared male parental visit duration to tadpole nurseries between captive (laboratory colony) and wild (Inselberg) Ranitomeya variabilis to assess whether the camera system captures comparable behavioral patterns across laboratory and field contexts.

Studies 1 & 2.

Behavioral scoring. Videos were processed and kept only when frogs entered the camera frame. The scorer marked entrance and exit times, defined as the moments at which the entire frog’s body was inside the focal canister or area. For benchmarking, observations were treated as discrete events. For the trigger-mode event-matching analysis, behavioral events were cross-matched between the two cameras using time-of-day from each camera’s internal clock, and detection was scored as TRUE if a given event appeared in a camera’s footage and FALSE otherwise. For the lab-versus-field comparison, a parental visit was defined as the entire body of an adult male entering the focal nursery, and visit duration was measured from entry to exit. For simplicity, frog attendance was measured as when the frog was present or absent for camera comparisons. In tadpole recordings, attendance was measured as the frog having the majority of its body (snout to vent) within the bromeliad or canister nursery measurements. For more qualitative behaviors not mentioned in the main text, we recorded the four qualitative and highly stereotyped behaviors described previously [25, 27]: egg care, tadpole transport, mating, pool visits (no tadpole), and pool visits (tadpole visits). Mating is when a calling male and female enter the canister together, huddling. Confirmation of mating is visually seen with the presence of an egg laid by the female. Egg care is the subsequent behavior of both sexes visiting the egg, often paired with huddling within a body’s length or on top of the egg itself. This is assumed to be parents hydrating via urination and/or guarding eggs. Tadpole transport is when a tadpole wriggles onto the adult’s back, which is visually confirmed with its body on the adult’s back. Tadpole transport concludes when the tadpole is physically off of the adult’s back. These observations are encoded in the Supplementary Data spreadsheet ‘CombinedVisits’ in the column ‘behaviors’.

Statistical analysis

Data processing was performed in RStudio (version 4.2.3). We used non-parametric tests throughout because Shapiro-Wilk tests indicated non-normal distributions of behavioral event durations and visit lengths and because of modest sample size. We used a Wilcoxon signed-rank test to compare paired event durations between Wyze v3 and GardePro in trigger mode (matched events recorded by both cameras), and a Wilcoxon rank-sum test to compare male parental visit durations between laboratory and field settings. P-values below 0.001 are reported as p < 0.001. Figures were prepared in ggplot2 [28] and assembled in Adobe Illustrator (Adobe Illustrator 2023).

Study 1: Results

Technical comparison of remote monitoring systems

Across the seven systems we compared (Table 1), home security cameras showed a distinct combination of features compared with conventional camera traps. They uniformly use pixel-change detection (enabling ectotherm capture), connect to local Wi-Fi rather than cellular networks, weigh 100–400 g, and cost $35 to $100 USD. Commercial trail cameras (Reconyx, Bushnell, GardePro) operated on AA batteries for long autonomous deployments and achieved better weather resistance (IP66) but relied on passive infrared (PIR) detection, which depends on body-heat differentials and performs poorly for small ectotherms, and cost $90–460 USD. The Wyze v3 and Tapo C120 specifically combined pixel-change detection, multi-spectral night vision (both near-IR at 940 nm and far-IR at 850 nm), IP65 weather rating, and prices of $35 to $40 USD, making them the most accessible candidates for ectotherm behavioral monitoring.

Laboratory benchmarking: In continuous mode, Wyze v3 detects more behavioral events than GardePro A3S

Over 96 h of simultaneous recording of the same focal area, Wyze v3 in continuous mode detected more behavioral observations than the GardePro A3S in trigger mode (n = 85 vs. 70 events, Fig 1C), although this difference was not statistically different between categories (χ2 = 1.45, df = 1, p = 0.23). The duration of behavior per observation also did not differ significantly between the two cameras (Wilcoxon rank-sum test, W = 3315.5, p = 0.20; Fig 1D). As Wyze records continuously, no event was lost to triggering failure. GardePro, by contrast, showed a triggering delay for most events: of the 68 GardePro events for which delays could be scored, 60 (88%, 95% CI: 78–95%) were missing an average 20.62 s of behavior per event compared with the continuous Wyze record (Fig 1E) and missing 324 s of behavior entirely. The cost of continuous recording was a very large fraction of empty footage, as only 1.8% of total Wyze recording time contained behavior (6,153 s with behavior vs. 333,900 s without; Fig 1F).

In trigger mode, Wyze records significantly longer behavioral segments per event

In trigger-only mode (75% motion sensitivity) and run alongside the GardePro, the Wyze produced fewer total triggers than the GardePro (n = 47 vs. n = 60; Fig 1G). On an event-matched basis, however, the Wyze v3 detected the same behavioral event more often (38/39 events, 97.4%) than the GardePro failed to (36/39 events, 97.4%; Fig 1H). For the subset of events that both cameras successfully detected, the Wyze recorded a significantly longer interval of behavior per event than the GardePro (Wilcoxon signed-rank test with continuity correction, V = 519, p < 0.001; Fig 1I).

Study 1: Discussion

Security cameras have previously been used to record behavior in frogs [29], arboreal snakes [30], fish [31], pigs [32], and chimpanzees [33]. Our laboratory results show that the Wyze v3, at approximately nine times lower cost than the GardePro, does not under-perform on any metric of behavioral capture relevant to this small ectotherm study, and outperforms the GardePro on event count (continuous mode), behavior lost per event (continuous mode), and behavior recorded per detected event (trigger mode). The single metric on which Wyze produced fewer numbers, total trigger count in trigger mode (Fig 1G), reflects fewer but longer recordings per detected event, not missed behavior, as the event-matching analysis (Fig 1H-I) shows Wyze does not miss more events than GardePro. Extending beyond the camera-comparison experiments, the Wyze system also captured detailed parental care behavior under laboratory conditions (Fig S1 in S1 File), demonstrating its utility for longitudinal studies, especially monitoring of behaviors that are rarely displayed under human observation.

Home security cameras are frequently designed to readily integrate with other software or hardware via the IFTTT (“If This, Then That”) platform [34], which enables automated workflows between programs from the moment a camera is triggered. Home security cameras can be paired with other programs, for example, with robotics for triggered behavioral manipulation ([35]), demonstrating broad cross-taxon adaptability for research. Our results establish that these cameras are suitable tools for laboratory research on small ectotherm behavior. In laboratory benchmarking, the home security camera matched or exceeded the behavioral performance of a commercial trail camera while remaining substantially less expensive.

The broader class of pixel-change home security cameras offers a viable alternative to commercial trail cameras for research and the market for such systems is rapidly expanding. Pixel-change detection is not unique to home security cameras; the more traditional Sony HDR-CX405, for instance, uses a CMOS (complementary metal-oxide-semiconductor) sensor, albeit with limited battery life. Both Wyze and Tapo cameras are lightweight, which is critical for field applications requiring long carries and extended deployments. The Tapo C120 (Table 1) is comparable in price to the Wyze v3 but has a higher IP rating and higher-resolution sensor, suggesting it may be more weather-robust and better suited to detecting small ectotherms. As the Tap C120 became commercially available after our study began, we could not benchmark it directly. However, the two-stage framework presented here, laboratory benchmarking against an established system, followed by field validation along independent hardware, sensitivity, and ecological-validity axes, provides a template for evaluating any such system. Reolink, Eufy, and Tapo/TP-Link all offer comparable pixel-change systems through globally distributed retailers, and Raspberry-Pi–based camera traps ([36]) allow further customization. Thus, our benchmarking design enables researchers to evaluate cost-effective options in advance of field work and identify the recording system best suited to their scientific question.

Combined with its lower cost, the laboratory benchmarking supports the use of the Wyze v3 over the GardePro A3S for small ectotherm recording despite Wyze’s lower weather-resistance rating (IP65 vs. IP66) and reliance on USB rather than AA power. The trigger-mode pattern of fewer but longer recordings per detected event is, for behavioral analysis, a feature rather than a deficit, as a single long recording per event preserves more behavioral context than several short fragments.

Study 2 Results

Hardware survival under tropical conditions

During the 1.5-month field deployment of 20 Wyze v3 units in French Guiana, 17 of 20 units remained functional to the end of the deployment despite daily rainfall of 0–30 mm and temperatures reaching 49°C at exposed rock surfaces in direct sunlight (Fig S3 in S1 File) with values up to 75 °C reported previously [23]. Three units required replacement: one camera cable was bitten by a rodent, one began producing pink-tinted recordings after one month, and one yielded little footage despite no visible damage.

Recording mode and sensitivity, once configured at the field station with Wi-Fi, were reliably preserved through offline operation. However, we identified the primary limitation to be power cycling. Each time the camera was disconnected and reconnected to a power source without Wi-Fi access, the internal timestamp reset to the factory default, and the camera reused folder names at every restart. This required SD-card replacement at every battery change to prevent file overwriting, effectively doubling the number of SD cards required relative to the number of cameras, and required filming a phone clock at startup to post-hoc reconstruct true recording times.

Sensitivity-versus-event-capture trade-off

Of sixteen tadpole-deposition events independently verified by daily manual inspection, six were captured on camera (37.5%). Despite these limitations, the camera system still captured a range of rare and difficult to observe behaviors under natural field conditions and at night, including heterospecific intrusion, predator encounters, intraspecific aggression, and an offspring-wiping behavior not previously documented in Ranitomeya variabilis (Fig S2 in S1 File, Supplemental Videos S1 Video–S4 Video).

Ecological validity of laboratory-measured behavior

We compared male parental visit duration to tadpole nurseries between captive (laboratory colony) and wild (Inselberg) R. variabilis to assess whether the camera system captures comparable behavioral patterns across laboratory and field contexts. Visit distributions overlapped substantially and the visit duration did not differ significantly between the two contexts (Kruskal-Wallis rank sum test, χ² = 0.42, p = 0.81; Fig 2C).

Study 2: Discussion

Differences among camera systems in both event detection and recording duration can affect which ectotherm behaviors are successfully captured and documented. For example, active infrared recordings at night may be detected by nocturnal species, but this has not been systematically measured thus far and may be highly variable [37]. In our study, small poison frogs use bromeliad leaves as tadpole nurseries, which cannot support the weight of a trail camera such as the GardePro, which would therefore require additional heavy mounting equipment, such as tripods, to be carried into the field. The Wyze v3’s lower missed-event rate and longer recorded duration per detected event, as established in the laboratory benchmarking, are directly relevant to monitoring poison frog behavior in the field. Many nursery visits are short in duration and could be missed entirely by a camera with a shorter trigger-to-recording window. Conversely, longer recording durations may improve the detection and documentation of more prolonged behaviors that could otherwise be truncated, such as heterospecific intrusions or the previously undescribed offspring-wiping event.

The hardware-survival rate (17 of 20 units over 1.5 months) was consistent with the camera’s intended outdoor home-security use. It exceeded our expectations given its substantially lower acquisition cost, as we had anticipated a higher failure rate under tropical field conditions. This supports its applicability to tropical deployments.

The choice between continuous and motion-triggered modes is the most consequential decision for deployment. At lowest sensitivity, the system effectively reduces false triggers and data volume but misses a substantial fraction of true behavioral events (capturing 6 of 16 independently verified deposition events). For quantitative behavioral counts under field conditions, continuous mode is therefore preferable to low-sensitivity triggering. Nonetheless, the capture of undescribed behavior illustrated that even low-sensitivity deployments can generate novel ecological and behavioral insights, demonstrating the broader utility of the system for cost-effective field-based behavioral studies in remote or extreme weather conditions. We therefore recommend continuous mode, potentially supplemented with post-hoc AI-assisted filtering, where quantitative behavioral inference is required and low-sensitivity motion-triggered mode where qualitative or complementary observations suffice. As recording mode and sensitivity, once configured at the field station with Wi-Fi, were reliably preserved through offline operation, these settings must be chosen before deployment rather than directly in the field.

Many behavioral studies on ectotherms are conducted under laboratory conditions without evaluating whether the recorded behaviors also occur under natural conditions. At the same time, researchers working in the Global South often rely primarily on field observations because extensive captive breeding facilities are not always available and because some species will not breed readily in captivity. The absence of a significant difference in male parental visit duration between captive and wild R. variabilis indicates that behavior monitored with the Wyze v3 under laboratory conditions reflects behavior expressed in the wild, rather than artifacts of captivity, supporting the application of this camera system across both captive and field study settings. Together, these results demonstrate that low-cost camera systems can bridge laboratory and field observations, validating each context against the other.

However, the Wyze system also showed several important limitations that should be considered before deployment. First, in many countries the newer Wyze v4 is more readily available than the v3 and offers improved image resolution and focus at a higher price and with the same constraints for offline use. Notably, full-length video recording and uninterrupted back-to-back event capture now require a subscription service, limiting the affordability of large-scale field research deployments. Second, the practical cost of continuous-mode or high-sensitivity operation in the field is data management as recordings are automatically organized into multiple nested folders that reuse identifiers across power cycles, and managing large datasets necessitates substantial time for sorting, renaming, and ensuring chronological accuracy across devices. Some ways to circumvent this challenge include the use of computer vision and deep learning for data extraction, which has been an active area of research in machine learning [38]. For example, vision language models, such as AVA, can triage meaningful events from false positives in long or continuous data streams [39]. Likewise, open-source AI models like MegaDetector [40] are specifically designed to detect animals in camera-traps with high accuracy for downstream processing. On balance, the Wyze v3 remained a cost-effective and field-capable choice for behavioral monitoring of ectotherms. Its principal limitations, including subscription dependence in newer models and the burden of large-scale data management, are increasingly tractable as both alternative home security systems and AI-assisted video analyses mature.

Conclusion

Our central motivation for this work is affordability and access to necessary scientific tools. Ectotherms constitute more than 99% of species diversity [41, 42], and much of that diversity is in the Global South, where research takes place within a context of historical and current colonialism and economic exploitation [43, 44] and where differences in science funding limit which questions can be asked [45]. Affordable remote monitoring is especially critical where ectotherm biodiversity is greatest and yet equipment budgets are most constrained. While the Wyze v3/v4 is currently shipped directly only within the United States and Canada and import costs can substantially increase the final price, we present it primarily as a proof of concept that home security cameras are viable low-cost alternatives to commercial trail cameras. Comparable systems are available through globally distributed retailers. An additional barrier to the broader use of home security cameras in research is language accessibility: the Wyze app interface and setup materials are primarily available in English. To reduce this limitation, we generated Spanish translations of the Wyze v3 app and quick-start guides, available via GitHub and Protocols.io (dx.doi.org/10.17504/protocols.io.dm6gp3p38vzp/v1), to help counteract the disadvantages created by the dominance of English in science [46]. We encourage researchers in the Global North to participate in efforts that improve the accessibility of scientific knowledge and research tools through collaboration and shared resources.

In summary, affordable home security cameras can match or surpass commercial trail cameras for image and video quality, detection methods, field of view, and motion-detection video length while being less expensive. Their application requires careful attention to sensitivity settings, power logistics, and data management, but the two-stage evaluation we present shows that they represent a substantial expansion of what is affordable for ectotherm behavior research, particularly in regions where the world’s ectotherm biodiversity is concentrated.

Supporting information

S1 File. Supplementary figures.

Supplementary figures and information on behavior and microhabitat.

(PDF)

pone.0360298.s001.pdf (5.8MB, pdf)
S1 Video. Footage of heterospecific intrusion.

(MP4)

Download video file (39MB, mp4)
S2 Video. Footage of predator encounter with spider.

(MP4)

Download video file (37.7MB, mp4)
S3 Video. Footage of intraspecific tadpole aggression.

(MP4)

Download video file (33.5MB, mp4)
S4 Video. Footage of offspring-wiping behavior in R. variabilis.

(MP4)

Download video file (31.8MB, mp4)
S2 File. Behavior Data.

This file contains parental behavior data.

(XLSX)

pone.0360298.s006.xlsx (37.2KB, xlsx)
S3 File. Behavior Data.

This file contains perch behavior data.

(XLSX)

pone.0360298.s007.xlsx (34.7KB, xlsx)
S4 File. Analysis Code.

This file contains R code used for analyses.

(R)

pone.0360298.s008.R (18KB, R)

Acknowledgments

We gratefully acknowledge Dr. Michael Hobbs, who brainstormed with BG in the early stages of her PhD how to best detect small frogs. We thank the Wyze community forum, where members and the company together brainstorm new ideas and troubleshoot current technology. We thank David Ramirez for their care of our domestic poison frog colony. This research was conducted at Stanford University, which is located on the ancestral and unceded land of the Muwekma Ohlone tribe. Our field work was conducted in the Nature Reserve “Les Nouragues” in French Guiana, that was founded on land ancestrally inhabited by the Amerindien Nouragues (“Norak”) tribe. We thank the Nouragues research field station (managed by CNRS) and are extremely grateful to the camp managers, and particularly Patrick Chatelet for his valuable technical, logistical and mental support throughout the field season.

Data Availability

Field videos, raw data, and as well as code for running visualizations can be found in supplementary materials. Benchmarking videos are available on DataDryad (10.5061/dryad.t1g1jwt78). Detailed instructions for setting up cameras can be found on Github (https://github.com/laurenoconnelllab/WyzeSetUps).

Funding Statement

This research was funded with grants from the National Institutes of Health (DP2HD102042, R01HD110514), the Rita Allen Foundation, the McKnight Foundation, Pew Charitable Trusts, and the New York Stem Cell Foundation to LAO. BCG was supported by an HHMI Gilliam Fellowship (GT15685) and a National Institutes of Health training grant (T32GM007276). MF was supported by the Erwin Schrödinger International Institute for Mathematics and Physics (J-4526B). ARL was supported by a Matthew Pecot Fellowship from the McKnight Foundation. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

  • 1.Sorge RE, Martin LJ, Isbester KA, Sotocinal SG, Rosen S, Tuttle AH, et al. Olfactory exposure to males, including men, causes stress and related analgesia in rodents. Nat Methods. 2014;11(6):629–32. doi: 10.1038/nmeth.2935 [DOI] [PubMed] [Google Scholar]
  • 2.Puścian A, Knapska E. Blueprints for measuring natural behavior. iScience. 2022;25(7):104635. doi: 10.1016/j.isci.2022.104635 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Martin PC, Bateson P. Measuring behaviour: an introductory guide. Cambridge, England: Cambridge University Press. 1986. [Google Scholar]
  • 4.Tape KD, Gustine DD. Capturing Migration Phenology of Terrestrial Wildlife Using Camera Traps. BioScience. 2014;64(2):117–24. doi: 10.1093/biosci/bit018 [DOI] [Google Scholar]
  • 5.Jachowski DS, Katzner T, Rodrigue JL, Ford WM. Monitoring landscape‐level distribution and migration phenology of raptors using a volunteer camera‐trap network. Wildl Soc Bull. 2015;39:553–63. [Google Scholar]
  • 6.Cserkész T, Kiss C, Sramkó G. Seasonal and diel activity patterns of small mammal guilds on the Pannonian steppe: a step towards a better understanding of the ecology of the endangered Hungarian birch mouse (Sicista trizona) (Sminthidae, Rodentia). Mamm Res. 2023;68:13–25. [Google Scholar]
  • 7.Buchholz R, Stamn J, Neha SA. Can camera traps be used to measure climate change induced alterations of the activity patterns of elusive terrestrial vertebrates?. Climate Change Ecology. 2021;2:100020. doi: 10.1016/j.ecochg.2021.100020 [DOI] [Google Scholar]
  • 8.Burton AC, Neilson E, Moreira D, Ladle A, Steenweg R, Fisher JT. Wildlife camera trapping: a review and recommendations for linking surveys to ecological processes. J Appl Ecol. 2015;52:675–85. [Google Scholar]
  • 9.Amber ED, Lipps GJ Jr, Peterman WE. Evaluation of the AHDriFT Camera Trap System to Survey for Small Mammals and Herpetofauna. Journal of Fish and Wildlife Management. 2021;12(1):197–207. doi: 10.3996/jfwm-20-016 [DOI] [Google Scholar]
  • 10.Hobbs MT, Brehme CS. An improved camera trap for amphibians, reptiles, small mammals, and large invertebrates. PLoS One. 2017;12(10):e0185026. doi: 10.1371/journal.pone.0185026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Welbourne DJ, Claridge AW, Paull DJ, Lambert A. How do passive infrared triggered camera traps operate and why does it matter? Breaking down common misconceptions. Remote Sens Ecol Conserv. 2016;2(2):77–83. doi: 10.1002/rse2.20 [DOI] [Google Scholar]
  • 12.Welbourne DJ, Paull DJ, Claridge AW, Ford F. A frontier in the use of camera traps: surveying terrestrial squamate assemblages. Remote Sens Ecol Conserv. 2017;3(3):133–45. doi: 10.1002/rse2.57 [DOI] [Google Scholar]
  • 13.Rovero F, Zimmermann F, Berzi D, Meek P. Which camera trap type and how many do I need? A review of camera features and study designs for a range of wildlife research applications. Hystrix-italian Journal of Mammalogy. 2013;24:148–56. [Google Scholar]
  • 14.Collen B, Ram M, Zamin T, McRae L. The tropical biodiversity data gap: addressing disparity in global monitoring. Trop Conserv Sci. 2008;1:75–88. [Google Scholar]
  • 15.Mugerwa B, Niedballa J, Planillo A, Sheil D, Kramer‐Schadt S, Wilting A. Global disparity of camera trap research allocation and defaunation risk of terrestrial mammals. Remote Sens Ecol Conserv. 2023;10(1):121–36. doi: 10.1002/rse2.360 [DOI] [Google Scholar]
  • 16.Whytock RC, Suijten T, van Deursen T, Świeżewski J, Mermiaghe H, Madamba N. Real-time alerts from AI-enabled camera traps using the Iridium satellite network: a case-study in Gabon, Central Africa. Ecology. 2021. [Google Scholar]
  • 17.Radke RJ, Andra S, Al-Kofahi O, Roysam B. Image change detection algorithms: a systematic survey. IEEE Trans Image Process. 2005;14(3):294–307. doi: 10.1109/tip.2004.838698 [DOI] [PubMed] [Google Scholar]
  • 18.Eddington VM, Nichols HK, Calistri-Yeh A, Young VKH, Kloepper LN. Graded alarm call behavior in wild fox squirrels (Sciurus niger). J Acoust Soc Am. 2024;155(2):1308–14. doi: 10.1121/10.0024771 [DOI] [PubMed] [Google Scholar]
  • 19.Hussey RK, LaPlante LH. Correlation between aggression and maintenance behaviors of captive Amphiprion ocellaris in a host anemone: A comparison between two variants. BIOS. 2024;95(1). doi: 10.1893/bios-d-22-00004 [DOI] [Google Scholar]
  • 20.Refinetti R. Diverse styles of running-wheel behavior in antelope ground squirrels. Behav Processes. 2020;177:104149. doi: 10.1016/j.beproc.2020.104149 [DOI] [PubMed] [Google Scholar]
  • 21.Shen JX, Xu ZM, Narins PM. Male antiphonal calls and phonotaxis evoked by female courtship calls in the large odorous frog (Odorrana graminea). J Comp Physiol A Neuroethol Sens Neural Behav Physiol. 2023;209:69–77. [DOI] [PubMed] [Google Scholar]
  • 22.Seki K, Tomiyasu M, Kuroda M, Ichimura M, Sato N, Zhu Y, et al. Temporal changes in behavior during the group spawning event of Pacific herring (Clupea pallasii). Sci Rep. 2025;15(1):11337. doi: 10.1038/s41598-025-95189-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Sarthou C, Kounda-Kiki C, Vaçulik A, Mora P, Ponge J-F. Successional patterns on tropical inselbergs: A case study on the Nouragues inselberg (French Guiana). Flora - Morphology, Distribution, Functional Ecology of Plants. 2009;204(5):396–407. doi: 10.1016/j.flora.2008.05.004 [DOI] [Google Scholar]
  • 24.Poelman EH, van Wijngaarden RPA, Raaijmakers CE. Amazon poison frogs (Ranitomeya amazonica) use different phytotelm characteristics to determine their suitability for egg and tadpole deposition. Evol Ecol. 2013;27:661–74. [Google Scholar]
  • 25.Brown JL, Morales V, Summers K. Divergence in parental care, habitat selection and larval life history between two species of Peruvian poison frogs: an experimental analysis. J Evol Biol. 2008;21(6):1534–43. doi: 10.1111/j.1420-9101.2008.01609.x [DOI] [PubMed] [Google Scholar]
  • 26.Poelman EH, Dicke M. Offering offspring as food to cannibals: oviposition strategies of Amazonian poison frogs (Dendrobates ventrimaculatus). Evol Ecol. 2007;21:215–27. [Google Scholar]
  • 27.Brown JL, Twomey E, Amézquita A, De Souza MB, Caldwell JP, Lötters S, et al. A taxonomic revision of the Neotropical poison frog genus Ranitomeya (Amphibia: Dendrobatidae). Zootaxa. 2011;3083(1). doi: 10.11646/zootaxa.3083.1.1 [DOI] [Google Scholar]
  • 28.Wickham H. Ggplot2. 2nd ed. Basel, Switzerland: Springer International Publishing. 2016. [Google Scholar]
  • 29.Gupta S, Marchetto PM, Bee MA. Customizable Recorder of Animal Kinesis (CRoAK): A multi-axis instrumented enclosure for measuring animal movements. HardwareX. 2020;8:e00116. doi: 10.1016/j.ohx.2020.e00116 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Stanca F, Soni S, Shaykevich D, Lorioux-Chevalier U, Kesseler P, Schlippe Justicia L. Underwater hunting in an arboreal snake: predation notes on amazonian frog larvae. Ecology and Evolution. 2026;16:e73376. [Google Scholar]
  • 31.McGaugh SE, Weaver S, Gilbertson EN, Garrett B, Rudeen ML, Grieb S. Evidence for rapid phenotypic and behavioural shifts in a recently established cavefish population. Biol J Linn Soc Lond. 2019. doi: 10.1093/biolinnean/blz162 [DOI] [Google Scholar]
  • 32.Netzley AH, Hunt RD, Franco-Arellano J, Arnold N, Vazquez AI, Munoz KA, et al. Multimodal characterization of Yucatan minipig behavior and physiology through maturation. Sci Rep. 2021;11(1):22688. doi: 10.1038/s41598-021-00782-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Havercamp K, Brindle M, Sommer V, Hirata S. Spontaneous nocturnal erections and masturbation in captive male chimpanzees (Pan troglodytes). Behav. 2022;159(12):1177–91. doi: 10.1163/1568539x-bja10166 [DOI] [Google Scholar]
  • 34.Mi X, Qian F, Zhang Y, Wang X. An empirical characterization of IFTTT: ecosystem, usage, and performance. In: Proceedings of the 2017 Internet Measurement Conference, 2017. 398–404.
  • 35.Chen TG, Goolsby BC, Bernal G, O’Connell LA, Cutkosky MR. Feed me: Robotic infiltration of poison frog families. arXiv. 2023. http://arxiv.org/abs/2305.14570
  • 36.Groffen J, Hoskin CJ. A portable Raspberry Pi-based camera set-up to record behaviours of frogs and other small animals under artificial or natural shelters in remote locations. Ecol Evol. 2024;14(3):e10877. doi: 10.1002/ece3.10877 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Meek PD, Ballard G-A, Fleming PJS, Schaefer M, Williams W, Falzon G. Camera traps can be heard and seen by animals. PLoS One. 2014;9(10):e110832. doi: 10.1371/journal.pone.0110832 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Vélez J, Castiblanco-Camacho PJ, Tabak MA, Chalmers C, Fergus P, Fieberg J. Choosing an appropriate platform and workflow for processing camera trap data using Artificial Intelligence. arXiv. 2022. 10.48550/arXiv.2202.02283 [DOI]
  • 39.Yan Y, Jiang S, Cao T, Yang Y, Yang Q, Shu Y, et al. AVA: Towards agentic video analytics with Vision Language Models. arXiv. 2025. 10.48550/arXiv.2505.00254 [DOI]
  • 40.Beery S, Morris D, Yang S. Efficient pipeline for camera trap image review. arXiv. 2019. 10.48550/arXiv.1907.06772 [DOI]
  • 41.Ohlberger J. Climate warming and ectotherm body size – from individual physiology to community ecology. Funct Ecol. 2013;27:991–1001. [Google Scholar]
  • 42.Wilson EO. The diversity of life. Cambridge, MA: Harvard University Press. 1992. [Google Scholar]
  • 43.Mahajan SL, Ojwang L, Ahmadia GN. The good, the bad, and the ugly: reflections on co-designing science for impact between the Global South and Global North. ICES Journal of Marine Science. 2022. [Google Scholar]
  • 44.Green T. North-South dynamics in academia. J Afr Cult Stud. 2019;31:280–3. [Google Scholar]
  • 45.Chankseliani M. Who funds the production of globally visible research in the Global South?. Scientometrics. 2023;128:783–801. [Google Scholar]
  • 46.Ramírez-Castañeda V. Disadvantages in preparing and publishing scientific papers caused by the dominance of the English language in science: The case of Colombian researchers in biological sciences. PLoS One. 2020;15(9):e0238372. doi: 10.1371/journal.pone.0238372 [DOI] [PMC free article] [PubMed] [Google Scholar]

Decision Letter 0

Patrick Stephens

25 Apr 2026

Dear Dr. O'Connell,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

I agree with the reviewers that the current organization of this manuscript is confusing.  I believe it would greatly benefit from following a standard intro, materials and methods, results, and discussion format.  The current paper also reads like two completely different papers woven together.  I would suggest that you solicit more input from the senior author to help revise the manuscript for improved organization and clarity.

Figure 1 contains the central results of the study.  The materials, methods and results related to this are the only part of the study that needs to be described in detail.  Most of the results described in figures 2-4 can be moved to the supplement, since the only point that they support is that the Wyze 3 can be used to capture a wide variety of behavioral data.   The details of the behavioral data captured don’t really matter in the context of the central question.

The materials, methods, and results related to the frog study system need to be much more clearly distinguished from the materials, methods, and results related to testing whether the Wyze 3 shows similar effectiveness to a more expensive system.  I would suggest creating a box that describes the frog study-system used for proof-of-concept analyses, to separate this text out from the material related to the central questions.

The results in figure 1 also need to be made much clearer.  For example, it is not immediately obvious that most of them differed significantly.  I would suggest that you  label the sets of observations that are significantly different using standard notation (e.g., an asterisk), with the test statistics and P-values from analyses also included in each section of the figure.  The other option would be to supplement the figure with a table listing each statistical analysis.  Regardless, readers shouldn’t have to dig into the text to figure out which of these sets of observations are significantly different at alpha.

You also need to explain much more directly whether and why these results support the use of the Wyze 3 over the GardePro.  It appears that the Wyze 3 often collected data that was significantly different from the GardePro.  However, in most of those cases it actually collected more data, with the exception of the observations in Figure 1G.    Using a standard format where (Results) you describe the results and then (Discussion) explain to us what conclusions you draw from the results would likely help with this.

One small point is that on the title page, the same symbol is used for the corresponding author and shared lead authorship.  Unless it is your intention to convey that there are three lead authors and three corresponding authors, you need to correct this.

Please submit your revised manuscript by Jun 09 2026 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

  • A letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

As the corresponding author, your ORCID iD is verified in the submission system and will appear in the published article. PLOS supports the use of ORCID, and we encourage all coauthors to register for an ORCID iD and use it as well. Please encourage your coauthors to verify their ORCID iD within the submission system before final acceptance, as unverified ORCID iDs will not appear in the published article. Only  the individual author can complete the verification step; PLOS staff cannot  verify ORCID iDs on behalf of authors.

We look forward to receiving your revised manuscript.

Kind regards,

Patrick R Stephens, Ph.D.

Academic Editor

PLOS One

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. To comply with PLOS One submissions requirements, in your Methods section, please provide additional information regarding the experiments involving animals and ensure you have included details on (1) methods of sacrifice, (2) methods of anesthesia and/or analgesia, and (3) efforts to alleviate suffering.

3. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

4. Thank you for stating the following financial disclosure:

“This research was funded with grants from National Institutes of Health (DP2HD102042), the Rita Allen Foundation, the McKnight Foundation, Pew Charitable Trusts, and the New York Stem Cell Foundation to LAO. BCG is supported by an HHMI Gilliam Fellowship (GT15685) and a National Institutes of Health Cellular Molecular Biology Training Grant (T32GM007276). MF is supported by an Erwin Schrödinger fellowship from the FWF (J-4526B). AL is supported by a Matthew Pecot Fellowship from the McKnight Foundation. LAO is a New York Stem Cell Foundation – Robertson Investigator.”

Please state what role the funders took in the study.  If the funders had no role, please state: "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

If this statement is not correct you must amend it as needed.

Please include this amended Role of Funder statement in your cover letter; we will change the online submission form on your behalf.

5. We note that Figures 1 and 2 in your submission contain copyrighted images. All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright.

We require you to either (1) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (2) remove the figures from your submission:

1. You may seek permission from the original copyright holder of Figures 1 and 2 to publish the content specifically under the CC BY 4.0 license.

We recommend that you contact the original copyright holder with the Content Permission Form (http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf) and the following text:

“I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.”

Please upload the completed Content Permission Form or other proof of granted permissions as an "Other" file with your submission.

In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].”

2. If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only.

6. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

7. If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously? -->?>

Reviewer #1: I Don't Know

Reviewer #2: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available??>

The PLOS Data policy

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English??>

Reviewer #1: Yes

Reviewer #2: Yes

**********

Reviewer #1: The introduction is very brief and does not describe the importance of your case study. Why did you choose this species and behavior? What were your predictions and aims for this study?

I understand that the PLOS ONE submission guidelines do not state that the materials and methods section needs to come before results and discussion, but these sections would benefit from a restructured organization that is smooth and easy to follow. I also understand that the results and discussions sections can be combined but I recommend being much more clear about what is a reported result and what is an interpretation for your disucssion. Everything from the beginning of the results and discussion section to line 122 on page 9 could be described as methods rather than results or discussion. Further, I recommend creating a new section after reporting your results where you interpret them, rather than immediately describing the importance of them.

I specifically focus on the paragraph starting on line 214 as an example of improving the overall structure of this study. In one paragraph, you describe your methods of measurement, report descriptive and analyzed statistics, and then intepret these results clamining that you have a novel finding on R. variabilis. While I appreciate the concise nature of such a paragraph, I lack the context as to why you set your methods this way, why you chose this test, and what the importance of this finding is to the species or goals of investigating this specices.

For your statistical reports, please clarify why you chose each non parametric test as a best fit for your data. Additionally, please italicize your p values and I strongly recommend reporting all values below 0.001 as <0.001. On Line 158, you report a p value left in exponential form, which makes it less accessible and clear.

Reviewer #2: Before making comments about the contents of the manuscript, I must indicate that I did not review the original version. The authors seem to have revised the manuscript in accordance with the suggestions made by the two reviewers. One issue I noticed was that a reviewer requested statistical tests for comparisons between two cameras; the authors said that they added tests, but I saw only one test in the whole article.

The article is interesting and is well written, although the style is heterogeneous. The article is partially a methodology study (of using simple cameras for the study of animal behavior) and partially a field report of frog behavior (Dendrobates tinctorius and Ranitomeya variabilis). For a methodology article, the description of frog behavior could be greatly shortened. Table 1 is the heart of the manuscript and provides the main information in the article. The provided information is valuable and should be of assistance to many field researchers.

As a minor comment, the authors should do a final reading of the manuscript to eliminate small errors. On page 4, line 89, the subject of the sentence is plural ("both home security cameras") but the verb is conjugated in the singular ("has"). On page 5, line 114, prices are said to range from upwards of $450 USD per camera to $35 USD; although the passage is not grammatically wrong, ranges are usually reported from lower to higher. On page 10, line 172, temperatures in the field are reported to range from 20.5°C to 50°C; this may be correct, but 50°C is a very high temperature; was this measured in the air (unlikely) or on the surface of an object under direct sunlight (more likely)?

**********

what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review?  For information about this choice, including consent withdrawal, please see our Privacy Policy

Reviewer #1: No

Reviewer #2: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

To ensure your figures meet our technical requirements, please review our figure guidelines: https://journals.plos.org/plosone/s/figures

You may also use PLOS’s free figure tool, NAAS, to help you prepare publication quality figures: https://journals.plos.org/plosone/s/figures#loc-tools-for-figure-preparation.

NAAS will assess whether your figures meet our technical requirements by comparing each figure against our figure specifications.

PLoS One. 2026 Oct 26;21(10):e0360298. doi: 10.1371/journal.pone.0360298.r003

Author response to Decision Letter 1

15 Jul 2026

Response to Review

We thank you and both reviewers for the constructive and detailed feedback. We have substantially revised the manuscript in line with the editorial and reviewer comments.

The principal change is structural. We agree the original framing made the manuscript feel heterogeneous and have reorganized the manuscript into the standard Introduction, Materials and Methods, Results, and Discussion format requested by the Editor and Reviewer #1. In doing so, we have refocused the manuscript explicitly as a two-stage methodology evaluation throughout: (i) laboratory benchmarking of the Wyze v3 against the GardePro A3S (now Fig 1, the central result), and (ii) field validation along three independent methodological axes (hardware survival under tropical conditions, sensitivity-versus-event-capture trade-off, and ecological validity of laboratory-measured behavior; now Fig 2). Behavioral observations not directly relevant to technical validation (sex differences in caregivers, canister preferences, opportunistic rare-behavior captures, Cruziohyla craspedopus observations) have been moved to Supporting Information. This reframing addresses the concern that the original manuscript read as "two completely different papers woven together".

We respond to each point below. Reviewer text is shown in italics; our responses follow each point. Page and line numbers refer to the revised manuscript with tracked changes.

Response to the Editor

"I agree with the reviewers that the current organization of this manuscript is confusing. I believe it would greatly benefit from following a standard intro, materials and methods, results, and discussion format. The current paper also reads like two completely different papers woven together."

Response: We agree on the structural point and have fully reorganized the manuscript into Introduction, Materials and Methods, Results, and Discussion. On the substantive concern about "two papers woven together," we have addressed this by reframing the field work as a methodological validation stage (along hardware, sensitivity, and ecological-validity axes) rather than as a separate behavioral case study. The Introduction now explicitly states that two complementary lines of evidence are required to establish a camera system for ectotherm behavioral research: laboratory benchmarking against an established system, and field validation under real-world conditions. The remainder of the manuscript follows this two-stage logic. The senior author has been closely involved in this revision.

"Figure 1 contains the central results of the study. The materials, methods and results related to this are the only part of the study that needs to be described in detail. Most of the results described in figures 2-4 can be moved to the supplement, since the only point that they support is that the Wyze 3 can be used to capture a wide variety of behavioral data. The details of the behavioral data captured don't really matter in the context of the central question."

Response: We have moved the behavioral-illustration content to the Supporting Information, in line with the Editor's recommendation. Original Fig 3 panels showing sex differences in caregivers and canister preferences are now S1 Fig (parental investment across offspring stage and environments). Original Fig 4 (opportunistic captures of rare behaviors) is now S2 Fig, and original Fig 5 (night recordings of Cruziohyla craspedopus) is now S4 Fig. We have retained two figures from the original manuscript in the main text in revised form, as they are essential to the field validation rather than illustrative of behavioral diversity: revised Fig 2 documents the field installation, the close-range refocus modification (hardware-survival evidence; this was original Fig 2), and the lab-versus-field comparison of male R. variabilis parental visit duration as the ecological-validity test (Fig 2C), simplified to two groups (captive males vs. wild males). Fig 1 (laboratory benchmarking) remains the primary result of the study.

"The materials, methods, and results related to the frog study system need to be much more clearly distinguished from the materials, methods, and results related to testing whether the Wyze 3 shows similar effectiveness to a more expensive system. I would suggest creating a box that describes the frog study-system used for proof-of-concept analyses, to separate this text out from the material related to the central questions."

Response: We have added a dedicated "Anuran study systems" subsection to Materials and Methods, which compactly describes the three frog species used: Dendrobates tinctorius as the focal species for laboratory benchmarking, Ranitomeya variabilis as the focal species for field validation, and Cruziohyla craspedopus for illustrative nocturnal observations in Supporting Information. The Materials and Methods of the main text is now organized around the two-stage evaluation: Camera systems compared, Laboratory benchmarking, Field validation (with the three validation criteria stated explicitly), Behavioral scoring, Statistical analysis, Camera setup, and Anuran study systems. This separates the methods directly required for the camera evaluation from the study-system descriptions, which are contained in the Anuran study systems subsection.

"The results in figure 1 also need to be made much clearer. For example, it is not immediately obvious that most of them differed significantly. I would suggest that you label the sets of observations that are significantly different using standard notation (e.g., an asterisk), with the test statistics and P-values from analyses also included in each section of the figure."

Response: We have revised Fig 1 accordingly. Each panel now displays the relevant significance marker directly above the compared groups (NS for not significant; *** for p < 0.001), with the test statistic and exact p-value printed in the panel (panel D: Wilcoxon rank-sum test, NS; panel I: Wilcoxon signed-rank test with continuity correction, V = 631, p < 0.001). The figure legend lists test details for each panel. Following Reviewer #1, p-values are reported in italics, and values below 0.001 are reported as p < 0.001 rather than in exponential form (the value originally reported as p = 2.54E-05).

"You also need to explain much more directly whether and why these results support the use of the Wyze 3 over the GardePro. It appears that the Wyze 3 often collected data that was significantly different from the GardePro. However, in most of those cases it actually collected more data, with the exception of the observations in Figure 1G. Using a standard format where (Results) you describe the results and then (Discussion) explain to us what conclusions you draw from the results would likely help with this."

Response: We have added a dedicated Discussion paragraph that explicitly addresses this. We state on which axes the Wyze outperforms GardePro (event count in continuous mode; behavior lost per event; behavior recorded per detected event in trigger mode) and explicitly resolve the apparent inconsistency in Fig 1G: although the Wyze produces fewer triggers in trigger mode, the event-matched analysis in Fig 1H to 1I shows that this corresponds to fewer-but-longer recordings, not to missed events. Combined with the roughly nine-fold lower unit cost, we interpret the result as supporting Wyze preference for laboratory and small-scale field studies of small ectotherms. We also state the two axes on which Wyze is not preferred (IP65 vs. IP66 weather rating; USB vs. AA power).

"One small point is that on the title page, the same symbol is used for the corresponding author and shared lead authorship."

Response: We corrected this issue. Shared first authorship is now denoted by the dagger symbol and the corresponding author by an asterisk, with separate footnotes.

Response to Journal Requirements

"Please ensure that your manuscript meets PLOS ONE's style requirements..."

Response: The manuscript has been reformatted to PLOS ONE templates for the title page, main body, and figure legends.

"In your Methods section, please provide additional information regarding the experiments involving animals and ensure you have included details on (1) methods of sacrifice, (2) methods of anesthesia and/or analgesia, and (3) efforts to alleviate suffering."

Response: We have added the following statement to the Camera setup and modifications methods subsection: "No animals were sacrificed or anesthetized in this study; cameras recorded freely behaving animals in their home enclosures and in natural habitats, and no procedures requiring anesthesia or analgesia were performed."

"We note that the grant information you provided in the 'Funding Information' and 'Financial Disclosure' sections do not match."

Response: This has been corrected on the submission portal.

"Please state what role the funders took in the study."

Response: We have added the requested statement: "The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

"We note that Figures 1 and 2 in your submission contain copyrighted images."

Response: In Figure 1, the Wyze copyrighted image was removed in the revision. In Figure 2, we took pictures of the equipment in our lab rather than using copyrighted images. See the original photo below:

"Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly."

Response: Supporting Information captions for S1 to S4 Figs and S1 to S6 Videos are now provided at the end of the manuscript. All in-text references have been updated to the new S-numbering.

Response to Reviewer #1

"The introduction is very brief and does not describe the importance of your case study. Why did you choose this species and behavior? What were your predictions and aims for this study?"

Response: We have expanded the Introduction. The final paragraph now states the three specific benchmarking questions we ask (event count, recording duration per event, missed-event rate) and the three field-validation criteria (hardware survival, sensitivity-versus-capture trade-off, ecological validity). Ranitomeya variabilis was selected for field validation because its parental behavior involves transient, hard-to-observe events occurring at small spatial scales, which is exactly the use case home security cameras would need to capture, and because its population at Inselberg, French Guiana, allowed paired laboratory and field deployment under extreme environmental conditions. This rationale is now stated explicitly in the Field validation subsection of Materials and Methods.

"I understand that the PLOS ONE submission guidelines do not state that the materials and methods section needs to come before results and discussion, but these sections would benefit from a restructured organization that is smooth and easy to follow."

Response: We agree. The manuscript has been restructured into separate Introduction, Materials and Methods, Results, and Discussion sections. Results are now strictly limited to what we observed; interpretation, mechanism, and recommendations are deferred to Discussion.

"Everything from the beginning of the results and discussion section to line 122 on page 9 could be described as methods rather than results or discussion."

Response: We agree and have changed this in the revised version. Camera class comparison criteria, benchmarking design (96 h periods, sensitivity settings, focal area), field-validation criteria, behavioral scoring rules, and the rationale for non-parametric tests are now all in Materials and Methods (subsections: Camera systems compared, Laboratory benchmarking, Field validation, Behavioral scoring, Statistical analysis). The Results section opens with the technical comparison (Table 1) and proceeds to the laboratory benchmarking and then the three field-validation results in sequence.

"I specifically focus on the paragraph starting on line 214 as an example of improving the overall structure of this study. In one paragraph, you describe your methods of measurement, report descriptive and analyzed statistics, and then interpret these results claiming that you have a novel finding on R. variabilis. While I appreciate the concise nature of such a paragraph, I lack the context as to why you set your methods this way, why you chose this test, and what the importance of this finding is to the species or goals of investigating this species."

Response: That paragraph has been decomposed across the new structure: (i) the methodological design (paired nursery and adjacent water pool; daily monitoring; lab vs. field comparison) is now in the Field validation subsection of Materials and Methods; (ii) the descriptive statistics and the Kruskal-Wallis rank sum test result are reported in Results under "Field validation 3: ecological validity of laboratory-measured behavior" and displayed in Fig 2C; (iii) the interpretation is reported in Discussion, now reframed as a methodological result (laboratory measurements generalize to natural conditions) rather than a behavioral finding about the species. The original behavioral-finding content (sex differences in caregivers; canister preferences) has been moved to S1 Fig as it is not central to the camera validation.

"For your statistical reports, please clarify why you chose each non parametric test as a best fit for your data."

Response: We have added a Statistical analysis subsection to Materials and Methods explaining the choice: non-parametric tests were used because Shapiro-Wilk tests indicated non-normality of the distributions of behavioral event durations and visit lengths, and because sample sizes were modest. For paired event-duration comparisons between Wyze and GardePro in trigger mode, the Wilcoxon signed-rank test with continuity correction was used because data were paired by event. For the lab-versus-field comparison of male parental visit duration, a Kruskal-Wallis rank sum test was used. The choice of test for each comparison is also restated at the point of use in Results and in the figure legends.

"Additionally, please italicize your p values and I strongly recommend reporting all values below 0.001 as <0.001. On Line 158, you report a p value left in exponential form, which makes it less accessible and clear."

Response: All p values are now italicized throughout the text, figure legends, and tables. The value previously reported as p = 2.54E-05 is now reported as p < 0.001. Test statistics and the abbreviation NS are italicized where used.

Response to Reviewer #2

"One issue I noticed was that a reviewer requested statistical tests for comparisons between two cameras; the authors said that they added tests, but I saw only one test in the whole article."

Response: Every comparison reported in Results is now accompanied by an explicit test, test statistic, p-value, and reference to the Methods subsection in which the test choice is justified. For Fig 1, the comparisons reported are: number of observations in continuous mode (descriptive, Fig 1C); observation length in continuous mode (Wilcoxon rank-sum test, W = 3315.5, NS, Fig 1D); seconds lost per recording in continuous mode (descriptive mean, Fig 1E); proportion of recording time with behavior (descriptive, Fig 1F); number of triggers in trigger mode (descriptive, Fig 1G); event-matched detection concordance (descriptive, Fig 1H); and event-matched behavior duration (Wilcoxon signed-rank test with continuity correction, V = 631, p < 0.001, Fig 1I). For the field validation ecological-validity test (Fig 2C), male parental visit duration was compared between captive and wild R. variabilis using a Kruskal-Wallis rank sum test (χ² = 0.42, p = 0.81, NS).

"The article is interesting and is well written, although the style is heterogeneous. The article is partially a methodology study (of using simple cameras for the study of animal behavior) and partially a field report of frog behavior (Dendrobates tinctorius and Ranitomeya variabilis). For a methodology article, the description of frog behavior could be greatly shortened."

Response: We agree the original framing was heterogeneous. We have restructured the manuscript as a two-stage methodology evaluation throughout: laboratory benchmarking (stage 1) followed by field validation (stage 2). The field validation is now explicitly framed as testing three methodological criteria (hardware survival, sensitivity-versus-capture trade-off, ecological validity of laboratory measurements) rather than as a behavioral study of R. variabilis. The behavioral observations not central to validation (sex differences in caregivers, canister preferences) have been moved to S1 Fig; opportunistic rare-behavior captures are now S2 Fig; and Cruziohyla craspedopus observations are now S4 Fig. The frog species used are now summarized compactly in an "Anuran study systems" subsection of Materials and Methods.

"Table 1 is the heart of the manuscript and provides the main information in the article."

Response: In the revised structure, Table 1 (technical comparison of seven camera systems) and Fig 1 (Wyze v3 vs. GardePro A3S benchmarking) together carry the central results. The Results section opens with Table 1 and proceeds directly to the empirical benchmarking presented in Fig 1, before turning to the field validation.

"On page 4, line 89, the subject of the sentence is plural ('both home security cameras') but the verb is conjugated in the singular ('has')."

Response: Corrected: "both home security cameras Wyze v3 and Tapo have multiple degrees of night vision capabilities..."

"On page 5, line 114, prices are said to range from upwards of $450 USD per camera to $35 USD; although the passage is not grammatically wrong, ranges are usually reported from lower to higher."

Response: Corrected. The sentence now reads "Camera prices ranged from $35 USD to upwards of $450 USD per camera."

"On page 10, line 172, temperatures in the field are reported to range from 20.5°C to 50°C; this may be correct, but 50°C is a very high temperature; was this measured in the air (unlikely) or on the surface of an object under direct sunlight (more likely)?"

Response: 50°C was a surface-temperature measurement. The temperature was recorded at exposed granite-rock surfaces in direct sunlight on the Inselberg of the Nouragues Nature Reserve. We refer to substrate temperature rather than air temperature because this measure is more biologically relevant for these frogs as they are in direct contact with the substrate, making substrate temperature a more accurate representation of the thermal conditions they experience than the temperature of the surrounding air.Even higher surface-temperatures (up to 75°C) were reported for this field site (Sarthou et al. 2009; ref 34). The text in both the Introduction (Field validation paragraph) and the Methods now clarifies this and we added a Supplementary Figure (Fig. S3) depicting the measurement location as well as the temperature profile recorded by the logger.

Attachment

Submitted filename: Response_to_Reviewers_v5.docx

pone.0360298.s010.docx (3.3MB, docx)

Decision Letter 1

Patrick Stephens

12 Aug 2026

Dear Dr. O'Connell,

Please submit your revised manuscript by Sep 26 2026 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

  • A letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

  • A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

  • An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

As the corresponding author, your ORCID iD is verified in the submission system and will appear in the published article. PLOS supports the use of ORCID, and we encourage all coauthors to register for an ORCID iD and use it as well. Please encourage your coauthors to verify their ORCID iD within the submission system before final acceptance, as unverified ORCID iDs will not appear in the published article. Only  the individual author can complete the verification step; PLOS staff cannot  verify ORCID iDs on behalf of authors.

We look forward to receiving your revised manuscript.

Kind regards,

Patrick R Stephens, Ph.D.

Academic Editor

PLOS One

Journal Requirements:

If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise.

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Additional Editor Comments:

This manuscript has been vastly improved compared to the original submission.  Unfortunately, the original reviewer that made the most suggestions on the original submission became unavailable to assess the revision.  However, we have secured another well qualified reviewer, that has a number of additional suggestions for clarity.  Since no new analyses will be required, it should be a relatively straightforward revision.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

Reviewer #2: All comments have been addressed

Reviewer #3: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions??>

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously? -->?>

Reviewer #2: Yes

Reviewer #3: No

**********

4. Have the authors made all data underlying the findings in their manuscript fully available??>

The PLOS Data policy

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English??>

Reviewer #2: Yes

Reviewer #3: No

**********

Reviewer #2: (No Response)

Reviewer #3: I think overall this is an excellent study and does valuable work advocating for understudied species and science accessibility. My issues with the manuscript are primarily in terms of organization. While there are marked improvements from the previous version, there are issues throughout where information appropriate for the introduction or methods section is not disclosed until the results or discussion.

I think the introduction would benefit from more detail to provide readers with necessary context. Some statements require citations (e.g. line 40 that Global South has much of ectotherm biodiversity). The paragraphs describing the differences between camera traps and home security cameras (lines 46-60) are lacking some potentially relevant details, such as the difference between how camera systems generally record daytime and nighttime videos. If there is some form of IR that can detect ectotherms, this should be mentioned in the introduction. While it makes sense that the introduction would focus on the differences between the two camera sources, I think the focal species and behaviors of interest should be described in the introduction, briefly, and in the context of what observational challenges would be directly improved by the results of this study. This information is present in Field validation subsection of methods, currently. The author's introduction should end with their hypotheses about how the two camera systems will perform in the lab and wild.

The methods section has discrete subheaders, which is great and a marked improvement in organization from the previous version. However, I think some of them would benefit from being expanded and reorganized. The authors might consider having a multistudy organization such that the camera systems comparison and laboratory study is reported in full (methods, results, discussion) before reporting on the wild study (with another methods, results, and discussion section), where the emphasis can remain on how the camera systems differed in terms of accurate event recording. The observations of C. craspedopus, unless they reveal something critical about Wayze' nighttime capacities not captured in the other two experiments, may be better off saved for a future manuscript. If this reorganization suggestion is pursued, anuran study systems can be separated and organized with the corresponding experiment. Otherwise, I think it would provide better narrative flow to place that section earlier in the methods, before behavioral coding. Environmental factors (brightness, humidity, etc) seem relevant in this case and could be included when describing the terrarium. Behavioral scoring would benefit from being accompanied by an ethogram of the behaviors studied. The amount of footage and total number of events (and total count per behavior) recorded by each camera system should be included. Finally, the "Statistical analysis" section should be the final section of methods and precede the results.

Given that no statistical analysis is necessary to describe the differences between the camera systems, Table 1 and lines 193-203 would fit better in the beginning of the methods section, under the "Camera systems compared" subheader (line 76), where the content of that table is first discussed. In the paragraph of lines 211-220, the authors argue that "Wyze v3 detects more behavioral events than GardePro A3S in continuous mode" however I don't see a p value to support this statement. If the loss of 21 seconds per recording led to statistically significant differences in numbers of events recorded, the test determining that should be reported. Along that line of thought, figure 1C and 1D would benefit from denoting significance level between camera types. Lines 228-237 are a lovely summary of the preceding results and are better suited as the first paragraph of the discussion.

Lines 257-306 return to the heterogenous organization and are a mixture of information appropriate for methods, results, and discussion. The authors should try to keep descriptions of field sites and procedure in the methods, limit the results to the outcomes of statistical analysis (or demographics of the dataset), and save the majority of interpretations or any mentions of alternative solutions to limitations for the discussion. The discussion in lines 388-404 is excellent and demonstrates the importance of this work. Information about camera systems differences and similarities might benefit from being described in the context of the specific frog behaviors observed by the study.

Typographical or grammatical errors:

line 35: "bth" rather than "both"

line 78: "...the suitability" it is more appropriate to use "their suitability" in this case

lines 108-113: This is a complex, multi-clause sentence. Consider breaking up into two or three sentences. Also, "air-to-surface temperature" is a bit unclear; is this the range of both air and surface temperatures? Consider reporting each individually.

line 195: The colon can be replaced by a period. As it stands, 194-197 is another run on sentence.

line 216-217: Consider breaking this up into two sentences (replaced "but" with "however")

line 286: difficult-to-observe; I think the hyphens are unnecessary

line 333: unnecessary comma

line 347: "identifying" should be "identify"

line 372: Again, consider replacing colon with period and making two separate sentences

**********

what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review?  For information about this choice, including consent withdrawal, please see our Privacy Policy

Reviewer #2: No

Reviewer #3: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

To ensure your figures meet our technical requirements, please review our figure guidelines: https://journals.plos.org/plosone/s/figures

You may also use PLOS’s free figure tool, NAAS, to help you prepare publication quality figures: https://journals.plos.org/plosone/s/figures#loc-tools-for-figure-preparation.

NAAS will assess whether your figures meet our technical requirements by comparing each figure against our figure specifications.

PLoS One. 2026 Oct 26;21(10):e0360298. doi: 10.1371/journal.pone.0360298.r005

Author response to Decision Letter 2

31 Aug 2026

Response to the reviewers

We thank the reviewers for their thoughtful and constructive comments, which have helped us to improve the manuscript. Below, we detail the changes made in response to each comment. Below, the reviewers’ comments are shown in black, followed by our responses in blue. For the reviewer’s convenience, all lines refer to the untracked version of the manuscript. We also found a small figure 1 error, which is now corrected (Fig 1H was accidentally labeled Fig1I, and Fig 1I was accidentally labeled Fig1J).

Reviewer #3:

I think overall this is an excellent study and does valuable work advocating for understudied species and science accessibility. My issues with the manuscript are primarily in terms of organization. While there are marked improvements from the previous version, there are issues throughout where information appropriate for the introduction or methods section is not disclosed until the results or discussion.

In the revised version, we have addressed this comment through the reorganization described point-by-point below.

I think the introduction would benefit from more detail to provide readers with necessary context. Some statements require citations (e.g. line 40 that Global South has much of ectotherm biodiversity).

We thank the reviewer for this feedback. For clarity, we have added detail and citations, specifically the following:

3. Martin PC, Bateson P. Measuring behaviour: An introductory guide. Cambridge, England: Cambridge University Press; 1986.

11. Welbourne DJ, Claridge AW, Paull DJ, Lambert A. How do passive infrared triggered camera traps operate and why does it matter? Breaking down common misconceptions. Remote Sens Ecol Conserv. 2016;2: 77–83.

12. Welbourne DJ, Paull DJ, Claridge AW, Ford F. A frontier in the use of camera traps: surveying terrestrial squamate assemblages. Remote Sens Ecol Conserv. 2017;3: 133–145.

13. Rovero F, Zimmermann F, Berzi D, Meek P. “Which camera trap type and how many do I need?” A review of camera features and study designs for a range of wildlife research applications. Hystrix-italian Journal of Mammalogy. 2013;24: 148–156.

14. Collen B, Ram M, Zamin T, McRae L. The tropical biodiversity data gap: Addressing disparity in global monitoring. Trop Conserv Sci. 2008;1: 75–88.

15. Mugerwa B, Niedballa J, Planillo A, Sheil D, Kramer-Schadt S, Wilting A. Global disparity of camera trap research allocation and defaunation risk of terrestrial mammals. Remote Sens Ecol Conserv. 2024;10: 121–136.

16. Whytock RC, Suijten T, van Deursen T, Świeżewski J, Mermiaghe H, Madamba N, et al. Real-time alerts from AI-enabled camera traps using the Iridium satellite network: a case-study in Gabon, Central Africa. Ecology. bioRxiv; 2021. Available: https://www.biorxiv.org/content/10.1101/2021.11.10.468078v1.full.pdf

17. Radke RJ, Andra S, Al-Kofahi O, Roysam B. Image change detection algorithms: a systematic survey. IEEE Trans Image Process. 2005;14: 294–307.

37. Meek PD, Ballard G-A, Fleming PJS, Schaefer M, Williams W, Falzon G. Camera traps can be heard and seen by animals. PLoS One. 2014;9: e110832.

The paragraphs describing the differences between camera traps and home security cameras (lines 46-60) are lacking some potentially relevant details, such as the difference between how camera systems generally record daytime and nighttime videos.

This has been expanded upon and clarified in lines 39-44 and 54-59.

If there is some form of IR that can detect ectotherms, this should be mentioned in the introduction.

There have been successful PIR recordings in mostly larger squamates such as crocodilians. In smaller squamates, PIR recording has been successfully done with a modified version of PIR called camera overhead augmented temperature (COAT) in Welbourne et al. 2013 (citation 12 above reviews this) which leverages the difference of squamate surface heat (similar to environment) compared to a cork background. Lines 41-43 hope to clarify this point.

While it makes sense that the introduction would focus on the differences between the two camera sources, I think the focal species and behaviors of interest should be described in the introduction, briefly, and in the context of what observational challenges would be directly improved by the results of this study. This information is present in Field validation subsection of methods, currently.

In response to this concern, a new paragraph was added to the end of the Introduction describing both study species (D. tinctorius and R. variabilis) and explaining their respective advantages for addressing the research question.

The author’s introduction should end with their hypotheses about how the two camera systems will perform in the lab and wild.

A new closing paragraph was added to the introduction stating the hypotheses of this work.

The methods section has discrete subheaders, which is great and a marked improvement in organization from the previous version. However, I think some of them would benefit from being expanded and reorganized.

We addressed this point via the restructuring described below.

The authors might consider having a multistudy organization such that the camera systems comparison and laboratory study is reported in full (methods, results, discussion) before reporting on the wild study (with another methods, results, and discussion section), where the emphasis can remain on how the camera systems differed in terms of accurate event recording.

We thank the reviewer for this helpful suggestion. Following this recommendation, we reorganized the manuscript using a multi-study structure. The manuscript is now divided into Study 1: Camera Comparison and Laboratory Benchmarking and Study 2: Field Validation, each with its own Results and Discussion sections. We retained a shared Materials and Methods section because both studies rely on common approaches for behavioral coding and statistical analyses. However, this section has been subdivided into methods common to both studies, methods and study system specific to Study 1, and methods and study system specific to Study 2.

The observations of C. craspedopus, unless they reveal something critical about Wayze’ nighttime capacities not captured in the other two experiments, may be better off saved for a future manuscript.

The C. craspedopus study system and all associated body text have been removed.

If this reorganization suggestion is pursued, anuran study systems can be separated and organized with the corresponding experiment. Otherwise, I think it would provide better narrative flow to place that section earlier in the methods, before behavioral coding.

We followed the reviewer’s advice and introduced D. tinctorius under Study 1 ("Anuran study system in the laboratory") and R. variabilis under Study 2 ("Anuran study system in the field").

Environmental factors (brightness, humidity, etc) seem relevant in this case and could be included when describing the terrarium.

We followed the reviewer’s advice, and this has been added under Study 1 (“Anuran study system in the laboratory”).

Behavioral scoring would benefit from being accompanied by an ethogram of the behaviors studied.

For simplicity, frog attendance was measured as when the frog was present or absent for camera comparisons. In tadpole recordings, attendance was measured as the frog having the majority of its body (snout to vent) within the bromeliad or canister nursery measurements. For more qualitative behaviors not mentioned in the main text, we recorded the four qualitative and highly stereotyped behaviors described previously in Brown et al., 2008: egg care, tadpole transport, mating, pool visits (no tadpole), and pool visits (tadpole visits). Mating is when a calling male and female enter the canister together, huddling. Confirmation of mating is visually seen with the presence of an egg laid by the female. Egg care is the subsequent behavior of both sexes visiting the egg, often paired with huddling within a body’s length or on top of the egg itself. This is assumed to be parents hydrating via urination and/or guarding eggs. Tadpole transport is when a tadpole wriggles onto the adult’s back, which is visually confirmed with its body on the adult’s back. Tadpole transport concludes when the tadpole is physically off of the adult’s back. These observations are encoded in the Supplementary Data spreadsheet ‘CombinedVisits’ in the column ‘behaviors’.

The amount of footage and total number of events (and total count per behavior) recorded by each camera system should be included.

We described this in passing in ‘Laboratory benchmarking: Wyze v3 detects more behavioral events than GardePro A3S in continuous mode’, but we agree with reviewers that this section would benefit from explicit descriptions of the footage and events recorded and therefore has been added.

Finally, the "Statistical analysis" section should be the final section of methods and precede the results.

Statistical analysis is now the last general Methods subsection (under "Studies 1 & 2"), immediately before the Results.

Given that no statistical analysis is necessary to describe the differences between the camera systems, Table 1 and lines 193-203 would fit better in the beginning of the methods section, under the "Camera systems compared" subheader (line 76), where the content of that table is first discussed.

We did not address this comment, as we respectfully disagree with the suggested change. Descriptive results are commonly reported without statistical analysis, for example in case studies, systematic reviews, or methodological and technical comparisons. The table summarizes the outcome of our systematic comparison of the camera systems and therefore represents a result of Study 1 rather than a description of the methods used. The comparison also informed the choice of camera model and will be useful to researchers selecting a camera system that best matches their study question. Accordingly, this comparison was considered one of the most important components of the study by other reviewers. We therefore consider Table 1 and the corresponding text better suited to the Results section.

In the paragraph of lines 211-220, the authors argue that "Wyze v3 detects more behavioral events than GardePro A3S in continuous mode" however I don’t see a p value to support this statement. If the loss of 21 seconds per recording led to statistically significant differences in numbers of events recorded, the test determining that should be reported.

This has been clarified to soften the language of this claim, given the count difference was statistically insignificant. However, it is difficult to run a statistical analysis for recording loss with Wyze in continuous mode vs GardePro, as our a priori expectations of data loss cannot be 50/50 with one camera recording 24/7 with 0 seconds of data loss. Thus, we added a comparison of the proportion of data lost (60/68 of events detected by GardePro experienced data loss; 88%, 95% CI 78-95%).

Along that line of thought, figure 1C and 1D would benefit from denoting significance level between camera types.

We thank the reviewer for this suggestion and have added “NS” to both panels to specify non significance.

Lines 228-237 are a lovely summary of the preceding results and are better suited as the first paragraph of the discussion.

This summary paragraph was merged with a sentence about the previous use of security cameras and now opens the Study 1 Discussion.

Lines 257-306 return to the heterogenous organization and are a mixture of information appropriate for methods, results, and discussion. The authors should try to keep descriptions of field sites and procedures in the methods, limit the results to the outcomes of statistical analysis (or demographics of the dataset), and save the majority of interpretations or any mentions of alternative solutions to limitations for the discussion.

We addressed the reviewers' concern by restructuring the manuscript. Specifically, field-site and procedural descriptions were moved into the Methods for Study 2, Results now report outcomes only. The interpretation of the results has been moved into the Study 2 Discussion, alongside the limitations.

The discussion in lines 388-404 is excellent and demonstrates the importance of this work.

We thank the reviewer and moved this general part of the discussion to a new “Conclusions” part with preserved verbatim.

Information about camera systems differences might benefit from being described in the context of the specific frog behaviors observed by the study.

We added an introductory paragraph to the discussion of the field study. In the revised version, it now reads:

“Differences among camera systems in both event detection and recording duration can affect which ectotherm behaviors are successfully captured and documented. Tadpole nurseries of small poison frogs, in this case bromeliad leaves, cannot support the weight of a trail camera such as the GardePro, which would therefore require additional heavy mounting equipment, such as tripods, to be carried into the field. The Wyze v3's lower missed-event rate and longer recorded duration per detected event, as established in the laboratory benchmarking, are directly relevant to monitoring poison frog behavior in the field. Many nursery visits are short in duration and could be missed entirely by a camera with a shorter trigger-to-recording window. Conversely, longer recording durations may improve the detection and documentation of more prolonged behaviors that could otherwise be truncated, such as heterospecific intrusions or the previously undescribed offspring-wiping event.”

Typographical or grammatical errors:

line 35: "bth" rather than "both"

line 78: "...the suitability" it is more appropriate to use "their suitability" in this case

lines 108-113: This is a complex, multi-clause sentence. Consider breaking up into two or three sentences. Also, "air-to-surface temperature" is a bit unclear; is this the range of both air and surface temperatures? Consider reporting each individually.

line 195: The colon can be replaced by a period. As it stands, 194-197 is another run on sentence.

line 216-217: Consider breaking this up into two sentences (replaced "but" with "however")

line 286: difficult-to-observe; I think the hyphens are unnecessary

line 333: unnecessary comma

line 347: "identifying" should be "identify"

line 372: Again, consider replacing colon with period and making two separate sentences

All typographical errors mentioned by reviewer 3 were corrected as suggested.

Reviewer #2: answered “Yes” to all assessment questions and provided no specific comments requiring a response.

Attachment

Submitted filename: Response to reviewer v6.docx

pone.0360298.s011.docx (145.5KB, docx)

Decision Letter 2

Patrick Stephens

24 Sep 2026

Home security cameras as a tool for behavior observations and science affordability

PONE-D-26-12504R2

Dear Dr. O'Connell,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. For questions related to billing, please contact billing support.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Patrick R Stephens, Ph.D.

Academic Editor

PLOS One

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

Reviewer #3: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions??>

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously? -->?>

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available??>

The PLOS Data policy

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English??>

Reviewer #3: Yes

**********

Reviewer #3: This looks great! I am pleased that the multistudy organization I suggested seems to have been a good choice. The authors' explanation for why Table 1 should remain in the result section makes sense, and I support the decision.

**********

what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review?  For information about this choice, including consent withdrawal, please see our Privacy Policy

Reviewer #3: No

**********

Acceptance letter

Patrick Stephens

PONE-D-26-12504R2

PLOS One

Dear Dr. O'Connell,

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS One. Congratulations! Your manuscript is now being handed over to our production team.

At this stage, our production department will prepare your paper for publication. This includes ensuring the following:

* All references, tables, and figures are properly cited

* All relevant supporting information is included in the manuscript submission,

* There are no issues that prevent the paper from being properly typeset

You will receive further instructions from the production team, including instructions on how to review your proof when it is ready. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few days to review your paper and let you know the next and final steps.

Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

You will receive an invoice from PLOS for your publication fee after your manuscript has reached the completed accept phase. If you receive an email requesting payment before acceptance or for any other service, this may be a phishing scheme. Learn how to identify phishing emails and protect your accounts at https://explore.plos.org/phishing.

If we can help with anything else, please email us at customercare@plos.org.

Thank you for submitting your work to PLOS One and supporting open access.

Kind regards,

PLOS One Editorial Office Staff

on behalf of

Dr. Patrick R Stephens

Academic Editor

PLOS One

Associated Data

    This section collects any data citations, data availability statements, or supplementary materials included in this article.

    Supplementary Materials

    S1 File. Supplementary figures.

    Supplementary figures and information on behavior and microhabitat.

    (PDF)

    pone.0360298.s001.pdf (5.8MB, pdf)
    S1 Video. Footage of heterospecific intrusion.

    (MP4)

    Download video file (39MB, mp4)
    S2 Video. Footage of predator encounter with spider.

    (MP4)

    Download video file (37.7MB, mp4)
    S3 Video. Footage of intraspecific tadpole aggression.

    (MP4)

    Download video file (33.5MB, mp4)
    S4 Video. Footage of offspring-wiping behavior in R. variabilis.

    (MP4)

    Download video file (31.8MB, mp4)
    S2 File. Behavior Data.

    This file contains parental behavior data.

    (XLSX)

    pone.0360298.s006.xlsx (37.2KB, xlsx)
    S3 File. Behavior Data.

    This file contains perch behavior data.

    (XLSX)

    pone.0360298.s007.xlsx (34.7KB, xlsx)
    S4 File. Analysis Code.

    This file contains R code used for analyses.

    (R)

    pone.0360298.s008.R (18KB, R)
    Attachment

    Submitted filename: PLOS_responsetoreview.pdf

    pone.0360298.s009.pdf (124.7KB, pdf)
    Attachment

    Submitted filename: Response_to_Reviewers_v5.docx

    pone.0360298.s010.docx (3.3MB, docx)
    Attachment

    Submitted filename: Response to reviewer v6.docx

    pone.0360298.s011.docx (145.5KB, docx)

    Data Availability Statement

    Field videos, raw data, and as well as code for running visualizations can be found in supplementary materials. Benchmarking videos are available on DataDryad (10.5061/dryad.t1g1jwt78). Detailed instructions for setting up cameras can be found on Github (https://github.com/laurenoconnelllab/WyzeSetUps).

    Articles from PLOS One are provided here courtesy of PLOS

    RESOURCES