Assessments of novel capture techniques are important to wildlife research. We used Comstock traps, a new live-capture technique, to capture North American river otters Lontra canadensis. We measured Comstock trap functionality in terms of river otter capture efficiency, furbearer capture efficiency, nonfurbearer capture efficiency, and malfunction rate. During 2014–2016, we captured 36 river otters (19 male, 17 female) in Comstock traps during 2,533 trap nights (1 capture/63 trap nights) at Crab Orchard National Wildlife Refuge in southern Illinois, USA. Eleven of 20 (55%) river otters assessed for capture-related injuries received an injury as a result of capture in a Comstock trap. The most common injury was claw loss (45%), followed by tooth fracture (25%) and lacerations (10%). The ease of setting Comstock traps and of releasing nontarget captures made them an appealing option for river otter live capture; however, two river otters died because of hypothermia, two died because of drowning, and one died because of traumatic injuries sustained during capture. Special care should be taken when selecting locations to set Comstock traps with regard to temperature and fluctuating water levels. Researchers attempting to live-capture river otters using this method would benefit by restricting their use to locations with predictable water levels and seasons with mild weather patterns.
Developing and adapting capture methods are important aspects of furbearer research (Kamler et al 2002; Kolbe et al. 2003; Koenen et al. 2005). Successful live-capture methods for furbearers are typically specialized to target a focal species (Kamler et al. 2002; McNew et al. 2007) while treating both target and nontarget species humanely (Olsen et al. 1986; Sikes and Gannon 2011). Optimizing furbearer capture techniques to maximize capture efficiency while reducing capture-related injuries has directed much of the research on trapping methods (Kreeger et al. 1990; Earle et al. 2003; Kolbe et al. 2003).
North American river otters Lontra canadensis (hereafter “river otters”) have traditionally been captured for research purposes using either foothold traps set in blind configurations (i.e., without bait) at latrines (Melquist and Hornocker 1983; Shirley et al. 1983; Serfass et al. 1996) or using cage-type traps such as box traps and Hancock cage traps (Northcott and Slade 1976; Shirley et al. 1983; Penak and Code 1987). However, river otters may avoid Hancock traps after their initial capture (Duffy et al. 1994), and river otters captured using Hancock traps can sustain substantial tooth damage (Blundell et al. 1999). Other studies have recommended foothold traps as a safe and efficient means to capture river otters (Melquist and Hornocker 1983; Serfass et al. 1996; Blundell et al. 1999). However, using foothold traps for river otters typically requires additional time in the field searching for suitable locations and prior knowledge and experience regarding how to set them effectively (Serfass et al. 1996; Blundell et al. 1999; Belfiore 2008).
We measured the functionality of a new design of box trap called a Comstock Custom Cage trap (Comstock Custom Cage LLC, Gansevoort, NY) (hereafter, “Comstock trap”) to capture river otters for a radio telemetry study (Rutter et al. 2018; Hanrahan et al. 2019). Comstock traps differ from other cage-type live traps in that they include a suspended wire trigger mechanism rather than a traditional foot treadle and spring-powered doors. These modifications allow the user to set the trap in a submerged or partially submerged setting and capture animals moving through the water. Our objectives were to measure injury rates, capture efficiency, malfunction rates, and nontarget capture rates of Comstock traps. To our knowledge, no other researchers have evaluated the functionality of this novel furbearer capture device.
Our study was conducted at Crab Orchard National Wildlife Refuge (hereafter, “the Refuge”) in southern Illinois. Southern Illinois is a temperate region typified by cold winters, wet springs, and hot and humid summers. Elevations ranged from 96 to 240 m. Annual average maximum temperatures from 1981 to 2010 ranged from 20.6°C to 8.8°C. Annual precipitation in the area averaged 125 cm, 19 cm of which occurred as snowfall (Arguez et al. 2010). The Refuge lies within the Big Muddy River drainage basin, which drains an area of 6,182 km2, and is one of the principal tributaries of the Mississippi River (Lewis 1955). The Refuge covers 17,761 ha in Williamson, Jackson, and Union counties and is managed by the U.S. Fish and Wildlife Service. Land cover on the Refuge consists of forests (56%), open water (20%), herbaceous (10%), croplands (10%), shrublands (2%), and developed land (2%; Frisk 2007). The Refuge is managed primarily for migratory waterfowl, native fishes, and several threatened and endangered species of plants and animals. Besides river otters, furbearers that occur on the Refuge include beaver Castor canadensis, bobcat Lynx rufus, coyote Canis latrans, gray fox Urocyon cinereoargenteus, long-tailed weasel Mustela frenata, mink Neovison vison, muskrat Ondatra zibethicus, Virginia opossum Didelphis virginiana, raccoon Procyon lotor, red fox Vulpes vulpes, and striped skunk Mephitis mephitis (Frisk 2007).
We captured river otters during two capture seasons: October 2014–January 2015 and October 2015–February 2016. During this period, juvenile river otters are typically weaned and capable of surviving independently (Hamilton and Eadie 1964). We tended between 10 and 40 Comstock traps daily between 0500 and 1000 hours and temporarily suspended trapping efforts when nightly temperatures were ≤ −12°C and precipitation events were forecast. We tested two sizes of Comstock traps during our study: the “12×12 Double Door trap” (smaller), and the “Double Door Beaver trap,” although we primarily used the former model. We set traps opportunistically in narrow stretches of streams, drainage ditches, and culverts where bottlenecks could be located. Where a bottleneck could not be located, we placed woody debris or rocks around the trap to further constrict the thoroughfare (Figures S1 and S3, Supplemental Material). We also set traps at overland travel routes between water bodies where river otter sign was identified (Figure S2, Supplemental Material). Previous studies have had success setting traps in “blind” configurations (i.e., without bait) to capture river otters entering and exiting latrine sites (Melquist and Hornocker 1979; Serfass et al. 1996; Blundell et al. 1999). We set traps at similar locations when they were found. However, the majority of our traps were set in bottlenecks along water courses where river otter sign was located. We did not utilize any bait or lure in either case. We camouflaged traps with grasses and other vegetation. We never set traps with more than 1/3 of the trap submerged, and we visited any traps that were partially submerged before other traps each day to minimize the time captured river otters spent in the water. We pinned open partially submerged traps with metal stakes whenever rain was forecast, in an attempt to ensure that no animals were captured when water levels rose. We equipped each trap with a 1- to 5-m cable anchored to a nearby object to protect against trap loss during flooding events. No additional efforts were taken to control human scent on traps before setting them in the field.
We visually estimated the weight of each river otter upon capture so that an adequate intramuscular injection of Telazol® (9 mg/kg body weight) could be delivered using a Pneu-Dart® CO2 pistol (Pneu-Dart Inc. Williamsport, PA). We estimated the age class of each captured river otter on the basis of tooth wear and discoloration using a scale developed by Belfiore (2008). We then transported anesthetized river otters to a nearby U.S. Fish and Wildlife Service facility to provide heat, electricity, and shelter to conduct injury assessments. We tagged each river otter with two individually numbered no. 1 Monel self-piercing ear tags (National Band and Tag Company, Newport, KY) and inserted passive integrated transponders (Biomark®, Boise, ID) under the skin between the scapulae to identify recaptures. We allowed river otters to recover within a protective box at the capture site. Capture and handling protocols were approved by the Southern Illinois University Animal Care and Use Committee (protocol #14–040).
We thoroughly inspected river otters for old and recent injuries to teeth and appendages during the second year of our study, following Blundell et al. (1999). Old dental injuries could be identified by discoloration at the broken area, smooth rounded edges at the fracture site, or a lack of recent gum damage when incisors were missing (Blundell et al. 1999); we did not include these in our analysis. We identified recent injuries to the appendages by edema or fresh lacerations, luxations, and fractures (Blundell et al. 1999). We scored recent injuries on the basis of a standardized scale of trauma developed by the International Organization for the Standardization of Traps (Olsen et al. 1986; Jotham and Phillips 1994; Table S1, Supplemental Material). We also measured the functionality of Comstock traps on the basis of capture efficiency (captures/trap nights) and malfunction rate (traps sprung/trap nights). Malfunction rate included any situation where a trap was rendered unable to capture a passing river otter for any reason including environmental conditions and flaws in trap functionality (e.g., set off by water flow, set off by floating debris, door caught open on floating debris, frozen open).
We live-captured 36 river otters (19 male, 17 female) in Comstock traps during 2,533 trap nights from October 2014 to February 2016 (Table 1). We live-captured 18 river otters and recaptured two from October 2014 to February 2015 during 1,014 trap nights. We live-captured 16 river otters and recaptured 4 from October 2015 to February 2016 during 1,519 trap nights. We captured 13 nontarget species, including, most commonly, raccoons, beavers, and red-eared sliders Trachemys elegans (Table 2). We documented four recapture events for river otters (all male), two of which occurred in traps set in the same location as their initial capture. We captured one male three times in traps set at two locations. On one occasion, we captured two river otters in the same trap at the same time. Five river otters (two males and three females) died as a result of capture complications (Table 3).
Eleven of 20 (55%) river otters inspected for injuries received some type of injury as a result of capture in a trap. Claw loss was the most common injury among river otters examined (45%), followed by tooth fracture (25%) and lacerations (10%). Mean ± standard error injury score for all animals examined was 32.1 ± 9.4, but was less when we excluded the four deaths that resulted from human error (13.7 ± 5.7).
We provide insight and recommendations regarding Comstock traps as a live-capture method for river otters. Given a recent focus on river otter conservation, especially in the Midwest where populations have rebounded from historic lows during the 20th century (Nielsen 2016; Rutter et al. 2018; Hanrahan et al. 2019; Holland et al. 2019), we believe that our findings are applicable to those considering alternative methods to live-capture river otters for research purposes.
Comstock traps possess several advantages over other commonly used methods used to capture river otters. In contrast to Hancock trap studies, we documented several river otter recaptures in Comstock traps. Both Duffy et al. (1994) and Blundell et al. (1999) reported that river otters became trap-shy after their initial captures in Hancock traps and thus did not report any recaptures. The powerful springs required to close the suitcase-style design of Hancock traps result in a particularly violent event when the trap is triggered. This event may be enough to reduce the potential for recapture in this style of trap. Furthermore, Comstock traps require less time to learn to operate and less time afield compared with alternative trapping methods. Where personnel learning to use foothold traps effectively must learn how to predict where an animal will potentially place its foot, personnel learning to use Comstock traps need only to place the trap in the bottleneck of a suspected thoroughfare. Additionally, in most cases setting foothold traps for river otters also entails finding an active latrine site (Serfass et al. 1996; Blundell et al. 1999; Belfiore 2008), which can require additional time spent afield. Comstock traps do not need to be set where a latrine is present and thus require less time afield. Additionally, this device can provide researchers the ability to use a wider range of potential set locations (e.g., crossovers and bottlenecks).
Although we captured many nontarget species using Comstock traps, those animals were relatively easy to release. Safety for both the researchers and the animals being captured was our foremost concern, and we found releasing nontarget captures from Comstock traps to be relatively stress free for both the researchers and animals. Of the eight mortalities of nontarget captures that occurred (Table 2), five resulted from human error associated with not pinning traps open before precipitation events. Three of the six wood ducks Aix sponsa captured in Comstock traps showed signs of predation through the trap mesh and likely would have survived otherwise.
Although we found Comstock trap efficiency to be comparable with others, our trapping effort exceeded those in the three most recently published studies (Serfass et al. 1996; Blundell et al. 1999; Belfiore 2008; Table S2, Supplemental Material). These studies adjusted their overall measures of capture efficiency to exclude periods where no captures occurred or when initial efforts to live-capture river otters proved unsuccessful because of flaws in equipment or initial trapping methods (Serfass et al. 1996; Blundell et al. 1999; Belfiore 2008). Thus, our study may more accurately represent the initial learning curve associated with efficiently live-capturing river otters using any method.
Our use of Comstock traps resulted in higher capture-related injury and mortality rates compared with other recent studies using alternative capture methods (Serfass et al. 1996; Blundell et al. 1999). Twenty-five percent of river otters we captured in Comstock traps sustained tooth damage. Although this rate was comparable with those reported by studies using other styles of trap (Serfass et al. 1996; Blundell et al. 1999), damage to the dentition of captured river otters is a major concern, as it is permanent and can seriously impede foraging (Biknevicius and Van Valkenburgh 1996). The rate of dentition injury we documented (25%) was lower than that of similar studies using Hancock traps. We hypothesize that this may be the result of differences in the mesh used to construct each type of trap. Hancock traps are typically constructed with galvanized steel “chain link” mesh material, which differs markedly from the wire mesh used to make Comstock traps. We used Comstock traps constructed with either 2.54-cm mesh or smaller 1.27-cm mesh. Although we did not compare injury rates between these trap types, anecdotal observations suggested that the smaller mesh size reduced space available to captured river otters' jaws, and thereby reduced the potential for dentition injuries. Additionally, some captured river otters received lacerations from what we assumed were the exposed Comstock trap trigger mechanisms (10% or 2 of 20 river otters). In one instance, this exposed trigger mechanism resulted in a major cutaneous laceration and an eye laceration. Future research should consider comparative analyses of injury rates between Comstock trap mesh sizes.
We report relatively high malfunction rates using Comstock traps, when compared with other methods. Our malfunction rate using Comstock traps (0.73) was higher than malfunction rates previously reported for Hancock (0.048) and foothold (0.065) traps (Blundell 1999), but this apparent difference might be a function of in-stream trap site selection (Figure S1). Although in-stream Comstock traps proved successful in capturing passing river otters, traps had to be reset frequently when triggered either by floating debris or nontarget captures. Although our higher malfunction rate should certainly be considered a disadvantage of this trapping method, it is worth noting that our trapping effort exceeded most others reported previously (Table S2).
The propensity for Comstock traps to capture nontarget species should be considered a disadvantage of this method. Frequent nontarget captures in Comstock traps likely influenced our river otter capture efficiency, as traps were unavailable to capture a passing river otter when a nontarget species was already captured. Nontarget captures were particularly common during our second field season (Table 2). Average temperatures in our study area were 3.5°C warmer during our second field season (Young et al. 2018), which likely had an influence on the activity levels of our most captured nontarget species (i.e., raccoons and red-eared sliders).
Foothold traps remain a relatively safe and effective method to live-capture river otters for those with the skill to use them, and they possess certain advantages over cage traps. Comstock traps can be particularly cumbersome to carry into remote study sites where automobile or boat access is limited and require a lot of space. Where one person walking to a remote field site would likely find it difficult to carry just one or two Comstock traps at a time, many foothold traps can be easily carried and take up far less space by comparison.
In terms of affordability, Comstock traps are also a particularly expensive option when compared with foothold traps. Comstock traps cost more than 10 times the recommended foothold trap (Sleepy Creek® Manufacturing #11 Long Spring = $13 vs. Comstock Custom Cage trap 12×12 Double Door trap = $158).
Mortality rates due to capture are important for researchers to consider. Mortality rates of river otters captured using Comstock traps placed in streams during our project were higher than expected, and possibly unacceptable to some researchers. Although we captured several adults when conditions were colder, two juveniles died when captured in Comstock traps because of hypothermia. We concluded that juvenile river otters likely had a harder time thermoregulating than larger-bodied adults when captured in partially submerged Comstock traps. We recommend establishing a more restrictive minimum water temperature cutoff for in-stream Comstock trap use of ≥ 1.6°C. However, we believe that future live-capture efforts could benefit by selectively using this device only when and where proper conditions exist. Restricting the use of Comstock traps to overland crossovers and trails leading to latrine sites during periods when overnight air temperatures are relatively warm (i.e., early fall in the Midwest) could reduce the risk of drowning and hypothermia for captured river otters.
The potential for semiaquatic mammals to drown in traps is an important consideration when choosing a capture method. Despite our best efforts to pin traps open before precipitation events and tend in-stream traps preferentially, two river otters drowned in traps. Conversion of hydrologic regimes for agricultural purposes can increase the magnitude and intensity of flooding events (Poff et al. 1997; Walser and Bart 2006), and many of the streams in our study area were channelized to control flooding and increase arable land. As a result, water levels in streams and ditches were difficult to predict (Poff et al. 1997), which made closing in-stream Comstock traps before rising water levels difficult. Future studies experiencing similar challenges could reduce the risk of drowning mortalities by reducing the number of traps maintained at one time and increasing the frequency of trap checks. Furthermore, simple modifications to the existing Comstock trap device might assist in reducing capture-related injuries to river otters. We believe replacing the acicular wire trigger design with a circular trigger wire coated in plastic or rubber would be beneficial. The same result might also be achieved by bending the ends of the trigger mechanism into loops.
Finally, we believe Comstock traps may have useful applications for researchers and managers attempting to live-capture other species of semiaquatic furbearing mammals. Live captures of beavers were particularly common during our project (Table 2). Although we did not compare the two box trap styles, Comstock traps seem to provide many of the benefits for beaver live-capture outlined by Koenen et al. (2005) when placed correctly without having to modify the original box trap design. Our experience shows that Comstock traps also could be useful for studies seeking to live-capture mink and muskrat if set in proper locations. In addition, we believe that this style of trap can provide a useful alternative for trappers using body-gripping traps to target beaver in areas where river otters occur. Incidental river otter captures during beaver-trapping activities occur frequently, even when following protocols to reduce incidental river otter harvest (Hamilton et al. 1998). Comstock traps can be used to live-capture both species in similar locations and can provide the opportunity to release river otters alive if set in a proper location where drowning and hypothermia are not a concern (e.g., overland crossovers and trails leading to latrine sites).
Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.
Table S1. Trauma scale used to assess the intensity of injuries sustained by river otters Lontra canadensis from October 2014 to February 2016 during live-capture efforts on Crab Orchard National Wildlife Refuge, Illinois (from Olsen et al. 1986; Jotham and Phillips 1994).
Found at DOI: https://doi.org/10.3996/092018-JFWM-083.S1 (36 KB DOCX).
Table S2. Capture efficiencies for river otter Lontra canadensis studies from 1976 to 2016 in North America in comparison with live-capture efforts from October 2014 to February 2016 on Crab Orchard National Wildlife Refuge, Illinois.
Found at DOI: https://doi.org/10.3996/092018-JFWM-083.S2 (36 KB DOCX).
Figure S1. Comstock trap used to capture river otters Lontra canadensis at a bottleneck along an in-stream travel corridor from October 2014 to February 2016 on Crab Orchard National Wildlife Refuge, Illinois.
Found at DOI: https://doi.org/10.3996/092018-JFWM-083.S3 (123 KB PDF).
Figure S2. Comstock trap used to capture river otters Lontra canadensis at an overland crossover between two water bodies from October 2014 to February 2016 on Crab Orchard National Wildlife Refuge, Illinois.
Found at DOI: https://doi.org/10.3996/092018-JFWM-083.S4 (417 KB PDF).
Figure S3. River otter Lontra canadensis captured during live-capture efforts from October 2014 to February 2016 in a Comstock trap set in an in-stream bottleneck on Crab Orchard National Wildlife Refuge, Illinois.
Found at DOI: https://doi.org/10.3996/092018-JFWM-083.S5 (108 KB PDF).
Reference S1. Frisk D. 2007. Crab Orchard National Wildlife Refuge comprehensive conservation plan. U.S. Fish and Wildlife Service, Crab Orchard National Wildlife Refuge, Marion, Illinois.
Found at DOI: https://doi.org/10.3996/092018-JFWM-083.S6 (10.05 MB PDF); also available at https://www.fws.gov/midwest/planning/craborchard/index.html.
Reference S2. Young AH, Knapp KR, Inamdar A, Rossow WB, Hankins W. 2018. The International Satellite Cloud Climatology Project, H-Series Climate Data Record Product. Earth System Science Data 10:583–593.
This research was funded by the Illinois Department of Natural Resources via Federal Aid in Wildlife Restoration Project W–135–R, the Department of Forestry, and the Cooperative Wildlife Research Laboratory at Southern Illinois University. B. Bluett, L. Hawk, T. Gettelman, C. Langan, J. Fort, M. Fisher, and H. Sullivan provided logistical and field support. We also thank the U.S. Fish and Wildlife Service staff at Crab Orchard National Wildlife Refuge for the use of their property and facilities. Additionally, we appreciate comments provided by reviewers and Associate Editor that strengthened earlier drafts of this manuscript.
Any use of trade, product, website, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government.
Citation: Rutter AU, Hanrahan AT, Nielsen CK, Schauber EM. 2020. Functionality of a new live-capture device for river otters. Journal of Fish and Wildlife Management 11(1):238–244; e1944-687X. https://doi.org/10.3996/092018-JFWM-083
The findings and conclusions in this article are those of the author(s) and do not necessarily represent the views of the U.S. Fish and Wildlife Service.