This paper describes a protocol that uses a remote video monitoring surveillance system to continuously monitor breeding colonies of ground-nesting waterbirds. The system includes five cameras monitoring individual nests and one camera monitoring the colony as a whole, and is powered by car batteries that are recharged via solar panels.
Many waterbird populations have faced declines over the last century, including the common tern (Sterna hirundo), a waterbird species with a widespread breeding distribution, that has been recently listed as endangered in some habitats of its range. Waterbird monitoring programs exist to track populations through time; however, some of the more intensive approaches require entering colonies and can be disruptive to nesting populations. This paper describes a protocol that utilizes a minimally invasive surveillance system to continuously monitor common tern nesting behavior in typical ground-nesting colonies. The video monitoring system utilizes wireless cameras focused on individual nests as well as over the colony as a whole, and allows for observation without entering the colony. The video system is powered with several 12 V car batteries that are continuously recharged using solar panels. Footage is recorded using a digital video recorder (DVR) connected to a hard drive, which can be replaced when full. The DVR may be placed outside of the colony to reduce disturbance. In this study, 3,624 h of footage recorded over 63 days in weather conditions ranging from 12.8 °C to 35.0 °C produced 3,006 h (83%) of usable behavioral data. The types of data retrieved from the recorded video can vary; we used it to detect external disturbances and measure nesting behavior during incubation. Although the protocol detailed here was designed for ground-nesting waterbirds, the principal system could easily be modified to accommodate alternative scenarios, such as colonial arboreal nesting species, making it widely applicable to a variety of research needs.
Common terns (Sterna hirundo, hereafter COTE), a waterbird species with a widespread breeding distribution, have become a flagship example of the need for conservation and monitoring programs1. Once harvested to near extirpation for the millinery trade, federal legislation in the 1900s enabled populations to rebound. However, declining population trends in the Chesapeake Bay have prompted increased concern over COTE, in addition to many other waterbirds2. COTE are currently listed as a Maryland state endangered species due to reductions in both breeding numbers and active breeding colonies3. Stressors including flooding and washouts of breeding sites4,5,6, anthropogenic disturbance, competition/predation with gulls7,8, and predation by great horned owls (Bubo virginianus) and red foxes (Vulpes vulpes)9,10, are believed to have contributed to current population declines; however, the relative contributions of individual stressors are not known. Understanding stressors associated with different stages of the breeding cycle, such as incubation, post-hatch, and fledging success are important but can be intensive and include frequent surveys that require entry into the nesting colony11. Such monitoring techniques can be disruptive to tern populations, and in some cases may result in nest abandonment and/or reductions in reproductive success12,13,14.
While the impact of researchers on common terns is well documented, intensive monitoring can impact a number of additional ground-nesting colonial species, such as short tailed shearwaters (Puffinus tenuirostris)15, common eiders (Somateria mollissima)16, black skimmers (Rynchops niger)17, and Fiordland crested penguins (Eudyptes pachyrhynchus)18. For instance, a study on short tailed shearwaters found that monitoring intensity had an inverse relationship on hatching success, and can exacerbate population declines. These examples illustrate the increasing need to reduce disturbance while maintaining comprehensive monitoring programs. With the video system outlined in this paper, we aimed to obtain information on nest attentiveness and observation of predators in a manner that would reduce the physical presence of humans within the colony.
Our study was located at the Paul S. Sarbanes Ecosystem Restoration Project at Poplar Island (38°46′01″N, 76°22′54″W, hereafter Poplar Island), one of the few known nesting sites for COTE in Maryland. Ongoing monitoring programs on Poplar Island have identified consistent nesting by COTE, albeit with variable levels of success depending on presence of avian or mammalian predators19,20. Due to these factors, Poplar Island was identified as an ideal location to conduct this study.
While the ability to monitor waterbird populations with video technology has clear benefits to the species under observation21,22, a number of technical considerations must be taken into account when implementing such an approach. For instance, video resolution must be sufficient to identify items of interest to the researcher, such as food items, nest markings, or colored leg bands for individual identification. Additionally, the physical components must be durable enough to withstand both weather events and wildlife interactions. Wireless security cameras were chosen due to their high definition picture quality, color display with wireless and infrared capabilities, outdoor durability, and overall cost effectiveness23.
The objective of this study was to design a video monitoring system that would allow for the remote observation of a ground-nesting colonial species while causing minimal disturbance to those individuals and the colony. This paper outlines the specific video system used to collect data.
1. Pre-Field Preparation of the Video Monitoring System
Note: This includes the steps necessary to prepare the solar panels, battery system, cameras and staking system for construction at the field site.
2. Construction of the Video Monitoring System in the Field
Note: This includes the steps necessary to secure the staking system, wire the battery system and solar panels, connect the DVR system and set up the cameras.
3. Review the Video Footage.
The implementation of this video monitoring protocol will result in continuous datasets of footage from five waterbird nests at close range and one set of footage of the entire colony from an elevated vantage point. A successful use of this system will minimize time where the footage is out of range or displaying a poor-quality image and will maximize time where the footage is of high quality (Figure 2; Figure 3). Cameras were installed on Poplar Island over the summer 2017 on nests with one egg because adult terns do not begin attending their nest regularly until a complete clutch (2-4 eggs) is laid24. Installation time averaged between 1-2 h, including both time inside the colony setting up the "Camera system" bin and time outside the colony setting up the "DVR system" bin. By installing the camera monitoring system prior to regular nest attendance, adult terns were given time to acclimate.
The fully functional electronic system resulted in the collection of 3,120 h of footage from 8 nests and 504 h of footage from one colony camera (Table 1). The nest count is more than five in this case because cameras with failed or successfully hatched nests were removed and reused for re-nesting attempts later in the season. Footage was recorded over 63 days in weather conditions ranging from 55 °F to 95 °F. Of these 3,624 h of footage, 3,006 h (83%) of footage contained usable data and 618 h (17%) of footage was out of range. Because the DVR system has a 2 TB capacity, the hard drive was replaced every 10 days. Signal interference was the primary cause of signal loss, though reduced charge on batteries and camera interference also contributed in small amounts. Footage obtained was sufficient to differentiate food items and nest markings, but insufficient to read numbers on the size 2 metal USGS bands. Behaviors such as feeding, incubation, nest maintenance, and preening were observed by adults on the nest level cameras and flocking and disturbance were observed by adults on the colony camera.
Total Footage (h) | Total Nests | Season Length (Days) | Lost Footage (h) | Usable Footage (h) | Percent of Footage Usable (%) | |
Ground Nest Cameras | 3120 | 8 | 63 | 508.52 | 2611.48 | 83.7 |
Colony Cameras | 504 | 1 | 63 | 109.82 | 394.18 | 78.2 |
Table 1: Overall video footage collection data for the 2017 nesting season. The following data was recorded during the 2017 common tern nesting season on Poplar Island in the Chesapeake Bay, Maryland.
Figure 1: Schematic of Parallel Battery Wiring System. A simplified diagram depicting the parallel wiring structure for the DVR system. The charge controller is depicted at the top with four car batteries forming a 2×2 pattern below. Please click here to view a larger version of this figure.
Figure 2: Visual representation of video footage quality. A. High quality imagery with no signal interference. B. Poor quality imagery with lines denoting signal interference. C. Out of range imagery denoting lost footage. A successful use of this system will maximize high quality footage and minimize both poor quality and out of range imagery outcomes. Note that B was captured from a separate trial location as opposed to A and C, which were captured on Poplar Island. Please click here to view a larger version of this figure.
Figure 3: Visual representation of video image from colony camera. Example of high quality imagery from the colony camera on Poplar Island. Please click here to view a larger version of this figure.
Monitoring waterbirds can be disruptive, and investigator disturbance while monitoring waterbirds has been linked to nest abandonment and decreases in reproductive success12,13,14. The protocol presented here offers a minimally invasive monitoring approach that allows researchers to establish and document the nesting behavior of ground-nesting waterbirds through continuous video footage.
Because this approach is minimally invasive, it requires a significant amount of initial labor to ensure the system operates smoothly when no researchers are present. One concern for researchers attempting to utilize this method is the maintenance of power required to charge the car batteries. The system presented here utilizes three solar panels for the DVR system and one solar panel for the camera system; however, maximum power output on solar panels varies based on latitude and time of year due to changes in day length and angle of the sun's path. We recommend testing all equipment prior to field use and verifying the correct angle for maximum sun exposure based on the location. There are several nonacademic user-friendly online sources to assist in determining optimal angle25,26. This test should occur in an area comparable to the intended site without impacting the target species. If the test run indicates that more power is necessary, the methods can be altered to add an additional solar panel to either the DVR or Camera systems.
During the initial establishment of this protocol, we encountered a number of challenges that required additional consideration and scrutiny. One such challenge was the maintenance of signal strength. The display monitor would consistently read either "out of range" or present data with a lot of noise for a number of cameras, demonstrating either a loss of signal or signal interference. Once we were confident the camera system was within range of the receivers and there were no objects obstructing the sightline, we began troubleshooting for alternative causes of signal interference. The biggest source of signal interference proved to be the placement of the receivers with respect to one another. According to the manufacturer, the receivers must be placed at least 4 in. away from one another to reduce interference27. If this is implemented and signal interference is still occurring, or there are too many objects obstructing the sightline, then receivers can be mounted to tall pipes in an effort to raise them above potential sources of signal obstruction. Alternatively, the wireless cameras could be replaced with wired cameras to ensure a strong connectivity. However, this would require a different cable system.
Additional considerations include weatherproofing all connections, ensuring the stability of the cameras and confirming receivers are properly secured. This system is constructed outside, and sand as well as moisture may enter the connections and cause corrosion. Protecting all connections with waterproof electrical tape will reduce the amount of corrosion and increase the longevity of the system. Along with sand and moisture, wildlife can also impact the system's longevity. During the establishment of this protocol, great horned owls were seen knocking cameras over multiple times, and COTE were often seen sitting on the colony camera, actions which can lead to data loss and damaged equipment. To reduce wildlife interactions, we recommend verifying the solidity of the camera placement in the ground prior to initializing the experiment. One possible adjustment could be the addition of rebar to the wooden camera stakes. Temperature may also have a significant impact on this system. Extreme temperatures can weaken the duct tape securing the receivers and may cause them to fall to the bottom of the container, resulting in lost footage. As a result, we recommend ensuring a strong receiver placement prior to deployment of the system. If extreme temperatures are common and/or a more secure attachment is needed, alternative attachment methods, such as a bracket system attaching each receiver to the bins, can be used. However, this would require more tools, labor and cost to establish.
This video monitoring protocol is limited in both location and monitoring potential. The staking system described here has been successfully tested on both sandy and semi-rocky terrain; however, it may be hard to set up on certain substrates such as rocky outcrops. This limits the potential use of the system to specific terrain. To bypass this problem, alternative staking methods suitable to the target terrain can be developed for the principal system, giving it a broader location potential. The monitoring potential is also limited by the finite scope of the infrared camera at night. Although this system utilizes a colony camera to see more of the colony than what is visible from the nest level, the infrared light does not fully extend across the colony and it is difficult to distinguish more than shapes outside the light range. As a result, the colony camera is very limited in its ability to provide insight into predation and overall colony patterns at night without the addition of external infrared lighting systems.
While the system described herein is for ground-nesting waterbirds, the principal system can be altered for other species, making it widely applicable to documenting the behavior of many target species. For example, for species that lay one egg, such as the Laysan albatross, researchers would need to install the system at nests at the time of nest construction to ensure the system is established prior to regular nest attendance. It is also critical to assess the target species tolerance towards the camera equipment before setting up the system, as this could limit effectiveness. This method enables researchers to collect data remotely with minimal disturbance and will allow researchers to continue documenting common tern nesting patterns and to expand remote monitoring efforts to many other species. It is our hope that this development can be utilized among the research and management communities to not only improve data collection, but also to reduce the impacts monitoring efforts have on the species we aim to protect.
The authors have nothing to disclose.
All data reported in this manuscript were collected in accordance with protocol approved by the Patuxent Wildlife Research Center Animal Care and Use Committee. This project was funded by the Maryland Department of Natural Resources and supported by the USGS Ecosystems Mission Area. Video production was funded by The Chesapeake Bay Trust and Friends of Patuxent. We would like to thank the U.S. Army Corps of Engineers, Maryland Environmental Service, and Maryland Department of Transportation Maryland Port Administration for general logistical support and allowing video filming on site. We would like to acknowledge Dr. Bill Bowerman and Dr. Daniel Gruner from the University of Maryland for their input into system design and implementation. We would also like to acknowledge Bill Schultz, Kaitlyn Reintsma and Katie DeVoss for their help in troubleshooting and in field set-up in summer 2017. Finally, we would to thank Michael Glow (internal review) and anonymous reviewers for their input. The use of trade, product, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.
Morningstar SS-20L-12V (2) | Morningstar Corporation | 3680192 | Charge controller |
Renogy 100 W 12V Panel (4) | Renogy | RNG-100D | Solar panel |
LOREX LW3211 (6) | Lorex | LW3211-2PK | Wireless camera with receivers |
Sawhorse (4) | HDX | SH106 | |
LOREX DV7082 | Lorex | DV7082W | 8ch 1080p HD DVR; Comes with computer mouse |
12V dry cell Absorbent Glass Mat (AGM) car batteries (6) | Optima | DS46B24R | |
TCT LCD color monitor | Kuman | X0013XAI51 | Mini display monitor |
22 in. display monitor | Dell | S2218H | For office |
18 gallon plastic bin (2) | Sterilite | 1446 | Plastic container |
Black copper insulated 10 AWG wire | Southwire | 22973257 | Black electrical wire |
Red copper insulated 10 AWG wire | Southwire | 37113803 | Red electrical wire |
3/8 in. ring terminals | Autocraft | 85417 | |
5/16 in. ring terminals | AutoCraft | 85445 | |
Winged wire connectors (red) | Commercial Electric | 775304 | Connector is large enough to accommodate 3 10AWG wires inside |
12V male DC adapter (2) | Avue | 162537 | |
Male DC 2.1 x 5.5 mm power plugs for CCTV (4) | WinBook | 231001 | |
Four port DC power splitters, 1 female to 4 | ClearView | PWRSPIDER4 | |
1.5 ft. wooden board (5) | Home Depot | 461443 | |
5 ft. wooden board | Vigoro | RC 85N | |
1/4 in. x 2 in. eye bolt (8) | Everbilt | 816721 | |
5/16 in. hex nuts (16) | Everbilt | 804886 | |
5/16 in. washers (16) | Everbilt | 807220 | |
SAE size 6 stainless steel clamps (8) | Everbilt | 670655E | |
60ft. BNC extension cables (6) | WinBook | 432377 | |
2 ft. x 4 ft. wooden plywood | Home Depot | 1502104 | Cut to 1 ft. x 2 ft. |
5 ft. metal rebar (8) | Weyerhaeuser | 35616 | |
Bungee cord (2) | HDX | 56128 | For securing lid |
15 ft. x 3/4 in. sticky back tape | Velcro | 239540 | |
Duct tape | Duck | 392875 | |
Permanent Marker | Sharpie | 35010 | |
1/4 in. x 400 ft. white diamond braid nylon rope | Everbilt | 72716 | |
Weatherproof electrical tape | Scotch | 6143-BA-10 | |
Schumacher 6A 12V automatic battery charger/ Carquest battery charger 8A | Schumacher/ Carquest | SP6/ CQ-80CR | Two possible car battery chargers |
6 in. nails (14) | Grip-Rite | 60HGC | |
18 Volt 1/2 in. Drill-Driver | Ryobi | P208B | Drill |
25 watt standard duty soldering iron | Weller | SP25NKUS | Soldering iron |
Leaded rosin core solder | Bernzomatic | 354123 | Solder |
Wire cutter | Stanley | 84-199 | |
Screwdriver | Husky | 146340142 | Came from 14 piece set of Phillips and flathead drivers |
15 in. aggressive tooth saw | Home Depot | 122SS159 | |
Rubber mallet | HDX | 31030 | |
Post driver | Everbilt | 901147EB |