Comparison of Small-Mammal Sampling Techniques in Tidal Salt Marshes of the Central Gulf Coast of Florida Melissa A. DeSa1, Christa L. Zweig1,*, H. Franklin Percival2, Wiley M. Kitchens2, and John W. Kasbohm3 Abstract - Use of traditional small-mammal sampling techniques is challenging in coastal salt marshes and forests. Various conditions particular to this type of environment affect the efficacy of sampling efforts. We compared various indirect (no physical capture) and direct (physical capture) techniques coupled with different baits to determine an effective method for sampling small mammals. In 1440 trap-nights, Sherman live traps caught significantly more rodents than Fitch traps (112 versus 16 respectively, Fisher’s exact test: P = 0.0103), but the relationship between bait types, which included oats, scratch feed, sunflower seeds and suet, was not as apparent. However, for practical reasons, scratch feed and sunflower seeds were best used in marsh conditions due to their water resistance and ease in clean-up. Both track and scat plates were unsuccessful due to water damage and destruction by nontarget animals. Cameras yielded limited success. Target small mammals were photographed 436 times, but of those, only 26 could be positively identified to species. However, with adjustments including improved lens quality and adjustable focus distance, this could become a useful tool for small-mammal sampling. Such improved remote cameras could offer a viable alternative when permits cannot be issued, field access is limited, or there are concerns about mortality of the target species. Introduction The central Gulf Coast of Florida is classified as a zero to low energy coastal environment where low relief, high tidal range, and minimal wave energy allow the persistence of extensive salt marshes (Montague and Wiegert 1990). Plant diversity is low due to the harsh physical environment that requires specialized salt and flooding adaptations (Stout 1984). Along the Gulf coast of Florida, Oryzomys palustris Harlan (Marsh Rice Rat), Sigmodon hispidus Say and Ord (Hispid Cotton Rat), and Peromyscus gossypinus Le Conte (Cotton Mouse) are common. The endangered Microtus pennsylvanicus dukecampbelli Woods, Post, and Kilpatrick (Florida Salt Marsh Vole) also inhabits this area, but has only been found in three locations (Doonan and Morgan 2002, Hotaling et al. 2010, Woods 1992). Small mammalian herbivores play important roles in salt marsh environments, shaping their habitat through foraging (Gedan et al. 2009) and serving as prey for numerous marsh predators, including raptors (Wolfe 1982). One of the most common small mammals occurring here is the Marsh Rice Rat, which impacts numerous trophic levels by feeding on plant parts, fungi, insects, mollusks, crustaceans, fish, bird eggs, and carrion (Negus et al. 1961, Wolfe 1982). 1Florida Cooperative Fish and Wildlife Research Unit, University of Florida, Box 110485, Building 810, Gainesville, fl32611. 2US Geological Survey, Florida Cooperative Fish and Wildlife Research Unit, University of Florida, Box 110485, Building 810, Gainesville, FL 32611. 3Sheldon –Hart Mountain National Wildlife Refuge Complex, 20995 Rabbit Hill Road, Box 111, Lakeview, OR 97630. *Corresponding author - czweig@ufl.edu. SOUTHEASTERN NATURALIST Gulf of Mexico Natural History and Oil Spill Impacts Special Series 2012 11(1):G17–G28 G18 Southeastern Naturalist Vol. 11, No. 1 Although abundant literature exists on various techniques available for smallmammal trapping (DeBondi et al. 2010, Eulinger and Burt 2011, Hayes et al. 1996, Rose 1994, Williams and Braun 1983), adapting these methodologies to tidal salt marshes is challenging. Daily tidal fluctuations submerge or flip traps, which can be a significant cause of mortality for trapped animals. They also present logistical challenges in accessing sites for researchers. In Florida, a trapping season limited to the cooler months (November through April) is advised by the US Fish and Wildlife Service due to increased chances of mortality from overheating in the hotter months. However, temperatures below freezing during the winter can also result in mortalities, especially in a wet environment. Furthermore, the Florida Fish and Wildlife Conservation Commission prohibits using peanut butter (a staple bait) and fruit, which attract Solenopsis wagneri Santschi (Red Imported Fire Ant) and can greatly influence mortality of trapped animals (Kraig et al. 2010, Masser and Grant 1986). Hotaling et al. (2010) found that attaching insulation foam boards to Sherman live traps with rubber bands, similar to that described by Wolfe (1985), was effi cient at keeping traps afloat. A survey flag or dowel placed through the foam and into the substrate was efficient in anchoring the trap assembly in place. It also prevented overturning by allowing the trap to rise and fall with tidal fluctuations (Fig. 1). Previous small-mammal work in this area had successfully captured the Salt Marsh Vole, Marsh Rice Rats, and Hispid Cotton Rats using Sherman traps (Woods 1982, Doonan and Morgan 2002) and the floating Sherman trap method (Hotaling et al. 2010). However, considerations such as rarity of Florida Salt Marsh Vole captures despite high effort, and the lack of trapping alternatives were incentives to explore other methods. The objective was to compare alternatives to the standard floating Sherman trap method by exploring both direct (live trapping, physically capturing the animal) and indirect (no physical capture) techniques. We sought an alternative that could increase detection probability of common small mammals inhabiting coastal marshes and forests, extend the study season beyond permissible live-trapping timelines, and provide another tool to Figure 1. Photo of floating Sherman trap set-up at Lower Suwannee National Wildlife Refuge during high tide. 2012 M.A. DeSa, C.L. Zweig, H. F. Percival, W.M. Kitchens, and J.W. Kasbohm G19 detect the Florida Salt Marsh Vole. Finding a comprehensive solution for effective trapping techniques is challenging due to differences in habitat and behaviors among the various small-mammal target species (Hayes et al. 1996, Slade et al. 1993, Weiner and Smith 1972). Given both logistical challenges and the target of multiple small-mammal species, indirect methods may be valuable tools to consider. Eliminating physical capture of the animal, indirect surveys provide a non-invasive alternative that can be used year-round without any risk of mortality. Furthermore, they eliminate the need for trapping permits, and reduce field efforts of setting and checking traps repeatedly. Remote wildlife cameras are one example of an indirect method, although relatively little work has focused on small mammals (De Bondi et al. 2010, King et al. 2007, Osterberg 1962). Most studies tend towards larger animals that are particularly rare or elusive species, or individuals that can be uniquely identified from photographs (Cutler and Swann 1999, Karanth and Nichols 1998, Trolle and Kéry 2003). Two other examples of indirect methods are track and scat plates, which rely on animals leaving paw prints and scat evidence that can later be identified for presence. These have been used more commonly for small mammals (Boonstra et al. 1992, Connors et al. 2005, Drennan et al. 1998, Hubbs et al. 2000, Justice 1961, Mabee 1998) compared to cameras, and are relatively easy to deploy. They do, however, require skilled identification of small prints, and scat that may look similar among related species. This small pilot study was performed to identify the most efficient sampling technique for capturing small mammals in the coastal habitats of the Gulf of Mexico. We compared Sherman and Fitch live traps in upland and adjacent Juncus roemerianus Scheele (Black Needlerush) marsh, three types of tracking plates, two types of scat plates, and wildlife cameras. Four types of baits were tested with the live traps. Methods Study area Our study area was located inside the Lower Suwannee National Wildlife Refuge (LSNWR), an area of over 21,000 ha protecting one of the largest, undeveloped, river-delta estuarine systems in the United States. The refuge is located along the northwestern Gulf coast of Florida (the “Big Bend region”), spanning Dixie and Levy counties (Fig. 2). The refuge contains the lower reaches of the historic Suwannee River, which flows from its headwaters in the Okeefenokee Swamp in Georgia, discharging into the Gulf of Mexico. Except for the Suwannee River Estuary, where freshwater discharge influences the surrounding environment, the typical Big Bend marsh ecosystem is characterized by broad stretches of salt marsh dominated by Black Needlerush, remnant stands of coastal forest and islands, meandering tidal creeks, and oyster bars (Raabe and Stumpf 1997). Slight variations in elevation, flooding, wave energy, and salinity determine presence and extent of Spartina alterniflora Loisel (Smooth Cordgrass), Black Needlerush, Distichlis spicata Greene (Saltgrass), and other high-marsh and salt-tolerant species (Montague and Weigert, 1990). The coastal forested islands are a mix of trees and shrubs, with variable G20 Southeastern Naturalist Vol. 11, No. 1 understory and groundcover species. Dominant tree species are Pinus spp. (various pine species), Quercus spp. (various oak species), Sabal palmetto (Walter) Lodd. ex Schult. & Schult. f. (Cabbage Palm), and Juniperus silicicola (Small) L.H. Bailey (Southern Red Cedar). This study took place both in upland coastal forested islands and adjacent marsh habitat (29.279°N, -83.073°W; Fig. 2). Figure 2. Location of the Lower Suwannee National Wildlife Refuge, flsampled for rodents using multiple direct and indirect survey techniques in 2010. Black star marks the location of our study area, a Black Needlerush marsh and its adjacent upland habitat. 2012 M.A. DeSa, C.L. Zweig, H. F. Percival, W.M. Kitchens, and J.W. Kasbohm G21 Indirect trapping techniques We used fifty 10- x 15-cm marine-grade plywood boards and fifty 10- x 15-cm foam boards as scat collection plates, which were baited with scratch feed (Flint River Mills, Inc., Bainbridge, GA) and deployed randomly throughout marsh and upland habitat. In each habitat, 25 foam and 25 wooden boards were used and placed 20–40 m apart. We checked scat plates approximately every three days depending on weather, throughout the month of May 2010. We used three types of tracking plates. The first consisted of a 1:2 parts carpenters chalk and rubbing alcohol solution, which was applied to a board of aluminum flashing with a garden-style sprayer similar to Drennan et al. (1998). The second consisted of powdered graphite suspended in an ethyl alcohol/ mineral oil solution (15%, 80%, and 5% respectively) applied to acetate transparency sheets with a foam brush similar to Connors et al. (2005). The third consisted of a copier paper strip with ink solution (powdered graphite and heavy mineral oil in a 1:2 ratio) soaked into felt pads stapled to either end of the paper similar to Mabee (1998). We clipped each track plate to aluminum flashing for stability and placed them inside a 7.5- x 10-cm PVC rain gutter down-spout cut to 30 cm length to protect plates from inclement weather. The housing was fitted with a 2.5-cm PVC feeding shelf attached inside the down spout similar to that used by Drennan et al. (1998). See Figure 3 for examples. We randomly placed 10 of each type of tracking plate (3 types, 30 plates total) in the marsh and checked them every three days for activity from 11 November– 18 December 2010 (1110 trap nights; 30 plates x 37 nights). Scuffed plates were replaced as required. We initially used scratch feed and sunflower seeds for bait; however, both were difficult to contain and thus damaged the plates. We switched to suet, which was easily contained in the feeding station. Game cameras were tested in two separate trials. In the first trial, eight motionsensor infrared cameras (Moultrie Game Spy® I-35, Model no. MSHF-DDC-50I) were tested for 24 trap nights (8 cameras x 3 nights) from 8–11 April 2010 in the marsh. They were set to capture a photo, 30-second video, time, date, moon phase, Figure 3. Examples of the tracking plates and their PVC housing: A) ink sheet with felt pads; B) carpenters chalk and rubbing alcohol solution plate with significant scuffing marks (similar appearance to the powdered graphite and mineral oil plate); and C) PVC housing fitted with a track plate, used to provide some weather protection. G22 Southeastern Naturalist Vol. 11, No. 1 temperature, and camera ID when triggered. Cameras were attached to fence posts 1.5 m above ground and angled downward to capture an area approximately 1 m2. This area contained two Sherman live traps with bedding material (cotton balls) baited with scratch feed. We attached traps to foam boards with rubber bands and staked them in placed using survey flags to keep traps afloat during high water events (Hotaling et al. 2010; Fig. 1). We opened traps at dusk and checked them at dawn for three consecutive nights. We released each captured small mammal immediately after identification. Cameras were checked daily for photos and to ensure functionality. The combination of cameras and Sherman live traps was intended to verify detection, should one method miss an animal. The second camera trial ran for 448 trap nights (8 cameras x 56 nights) from 18 October–12 December 2010. For the second trial, we set cameras in upland areas only, to reduce false triggers from moving vegetation. We fastened cameras to trees or logs close to the forest floor, with the lens aiming at a small pile of scratch feed mixed with sunflower seeds on the ground. As images were reviewed, we adjusted camera positions and angles to increase clarity of image. Scratch feed and sunflower seeds were alternately used for bait each week. No video settings were used in this trial to conserve memory. Cameras were left running and checked periodically to ensure operability throughout the 56-night period. Direct trapping techniques Two sessions of 4 trap nights each were performed from 11–15 November and 4–8 December 2010 at three sites for a total of 1440 trap nights (2 sessions x 4 nights x 60 traps x 3 sites). The three sites included two upland forested areas and one Black Needlerush marsh. Each site had thirty 22.9- x 7.6- x 8.9-cm Sherman and thirty 38- x 6- x 6-cm Fitch traps arranged in sets side by side and spaced approximately 3 m between sets. Fitch traps are constructed of hardware cloth and aluminum soup cans (Rose 1994), providing more ventilation and an alternative to the metal box-style Sherman live trap. One upland site consisted of a drift fence 15 m long x 1.5 m tall, installed just above the high tide mark and lined on either side (i.e., marsh side and upland side) with trap sets. The other upland site consisted of a trapping grid with trap sets arranged in three rows of 10 (27- x 6-m grid area). The marsh site was a linear transect with no drift fence, placed in monospecific Black Needlerush, using the floating trap design (Fig. 1). Four bait types were tested. The first trapping session used oats and scratch feed, the second used suet and sunflower seed. Along trap lines, each set alternated baits. Fisher’s exact test was performed to test for nonrandom associations between two categorical variables with small sample sizes (i.e., traps and baits). Results Indirect trapping techniques Both foam and wooden scat plates contained small-mammal droppings, but the plywood did not float consistently and became saturated with water, damaging or washing away many samples. Although the foam boards floated well, the samples were often wet, affecting the quality of the scat sample. The only 2012 M.A. DeSa, C.L. Zweig, H. F. Percival, W.M. Kitchens, and J.W. Kasbohm G23 tracking plate that provided a distinguishable print was from a single ink sheet that produced one small-mammal print. Despite the protection provided by rain gutters, many plates were affected by heavy rains. All plates were marked by non-target animals attracted to the suet, damaging plates and making any tracks difficult to discern. The first camera session yielded no small-mammal pictures; the video card was primarily occupied by photos and video of false triggers. During the second session, a total of 2544 photos were taken, which included 1881 false triggers (73.9% of all photos) with no animals in the frame. Of the 663 animal photos (Table 1), 34% were non-targets including Dasypus novemcinctus L. (Nine-Banded Armadillo); Lynx rufus floridanus Rafinesque (Florida Bobcat); Glaucomys volans L. (Southern Flying Squirrel); Urocyon cinereoargenteus Schreber (Grey Fox); Didelphis virginiana Kerr (Virginia Opossum); Procyon lotor L. (Northern Raccoon); and various birds, crabs, and insects. Of the 436 (66%) photos containing target small mammals, only 26 (5.9%) were positively identified, many of which were large rats, such as Rattus norvegicus Berkenhout (Norway Rat) or Neotoma floridana Ord (Eastern Woodrat). Cotton Mice and Marsh Rice Rats were photographed, but most photos were impossible to verify. Picture quality and angle were not ideal, and many photos were unfocused, or only a portion of the animal was visible, making identification impossible. No Florida Salt Marsh Voles were identified in any photos. Direct trapping techniques During the live-trapping pilot study, no animals were captured in the upland habitats containing the grid and drift-fence arrangements. In the marsh habitat, Marsh Rice Rats were captured more frequently than Hispid Cotton Rats (118 times versus 10, respectively). No Florida Salt Marsh Voles were captured. Grouping both species, the Fisher’s exact test indicated that there is a significant Table 1. Target (rodent) and non-target animals photographed using wildlife cameras across 448 trap nights (18 October–12 December 2010) in upland habitats at the Lower Suwannee National Wildlife Refuge, FL. Common name Number of Photos Cotton Mouse 10 Hispid Cotton Rat 1 Marsh Rice Rat 15 Unknown Rodent 410 Florida Bobcat 4 Grey Fox 4 Nine-banded Armadillo 3 Northern Raccoon 37 Southern Flying Squirrel 3 Virginia Opossum 7 Birds (mostly Sora) 105 Unknown snake 1 Fiddler crabs 53 Insects 10 Total 663 G24 Southeastern Naturalist Vol. 11, No. 1 association between trap and bait type (P = 0.0103; Table 2). Fisher’s exact test does not indicate that the significant difference between trap and bait types are the same, but that there is a significant association. For trap type, interpretation is clear: Sherman traps caught more individuals compared to Fitch traps (112 or 46.7% capture success versus 16 or 6.7% capture success). However, the difference between baits is less apparent—only scratch captures in Fitch traps differs much from the expected. Discussion Various small-mammal sampling techniques were tested in tidal forest and marsh environments in the Big Bend region of north Florida. We sought an effective method that might extend the study season, increase detection probability, and provide another tool to detect the endangered Florida Salt Marsh Vole. Experiments with indirect trapping techniques that included track plates, scat plates, and game cameras were mostly unsuccessful. Of these, cameras were the most promising. Direct trapping techniques were most effective. Both foam and wooden scat plates contained droppings, but the method was generally unreliable due to inconsistent floatation of the plywood, and scat being washed away, blown off, or saturated for both plate types. Good quality, intact samples are required for field identification, particularly when distinguishing among related animals where error rates can be as high as 50% (Halfpenny 1986). Had better samples been available, we would have compared to known samples from trapped animals. An alternative identification technique is DNA analysis (Foran et al. 1997, McKelvey et al. 2006, Reed et al. 1997). This was beyond the scope of our project, but may be an option for other researchers acquiring higher quality samples. However, the cost of running the tests should be weighed against the costs of other methods available to sample small mammals. Tracking plates also were ineffective. The ink plates produced only one unidentifiable print which was difficult to see among debris such as vegetation, sand, and other animals that marked up the plate. The mark itself was also unclear. The other plate types were commonly damaged by crabs which were Table 2. Counts of trapped small mammals using various trap and bait combinations sampled from 11–5 November 2010 and 4–8 December 2010 in Black Needlerush marsh habitat at the Lower Suwannee National Wildlife Refuge, fland expected contingency table for Fisher’s exact test. The Fisher’s exact test indicated there were significant associations between bait and trap types (P = 0.0103). Oats Scratch Seed Suet Total Actual captures Fitch 3 10 2 1 16 Sherman 45 23 27 17 112 Total 48 33 29 18 128 Expected contingency table Fitch 6 4 4 2 16 Sherman 42 29 25 16 112 Total 48 33 29 18 128 2012 M.A. DeSa, C.L. Zweig, H. F. Percival, W.M. Kitchens, and J.W. Kasbohm G25 observed on the plates consuming suet. Northern Raccoons and Nine-banded Armadillos may have also contributed to the damage, which was enough to remove any evidence of small-mammal tracks. Suet was the only bait that did not damage the plates, and thus it was used almost exclusively. However, it appeared to be a major attractant to non-target animals, causing significant damage. The rain gutters, intended to provide weather protection, were not sufficient during heavy rains. We found many plates to be affected by water. Both scat and track plates have been useful in other studies (Barrett 1983, Cain et al. 2006, Connors et al. 2005) and provide a viable sampling technique in other environments. However, it appears that in coastal salt marsh and forests that experience tides, copious rain, and an abundance of non-target wildlife, scat and track plates are ineffective for sampling rodents. A few other adjustments might be used to improve efficacy of track plates, including increased rain protection for the ethyl alcohol/mineral oil plates, which were the most water resistant. An alternative compact bait could be tried, such as a non-toxic rodent bait block, which may be less palatable to crabs and other non-target animals compared to suet. The cameras were more successful compared to track and scat plates. With continued testing and modification, they might provide a valuable alternative to live trapping during the off-trapping season in Florida (May through October). Although some photos were positively identified as target small mammals, these constituted a very small percentage, and most small mammals were not identifiable. Many were either blurry, captured the animal at an angle that made it difficult to confidently identify, or only showed a segment of the animal. De Bondi et al. (2010) also found small-mammal identification challenging and suggested clearing the area to expose bare ground, thereby increasing the contrast between the animal and its background. This allows for the ratio between body and tail to be more easily determined, which helps with identification. It might also be useful to place an object of known size or a ruler within the view of the camera for scale (Osterberg 1962). Alternative cameras that produce higher quality images, and have options to adjust response time, shutter speed, lens quality, and adjustable focus distance should be tested. RECONYX® brand cameras are high quality wildlife cameras that might be considered (RECONYX®, Inc., Holmen, WI). Higher quality cameras are coupled with a larger initial expense that may not be feasible for some research budgets. Some argue that if all equipment must be initially purchased, then cameras are less expensive compared to live-traps, and are also less labor intensive to maintain (De Bondi et al. 2010, Srbek and Chiarello 2005). However, convenience and cost may be weighed against the objectives of the study and the potential problems inherent in “capturing” small mammals by photo. Observing small mammals with cameras is a fairly novel use, as they are typically used for medium- to large-sized animals (Karanth and Nichols 1998, Srbek and Chiarello 2005, Trolle and Kéry 2003, Vine et al. 2009). Recent attempts to use camera traps to estimate occupancy and detection probabilities for small and large mammals (De Bondi et al. 2010) may take more time to be fully integrated and utilized (O’Connell and Bailey 2010). The live-trapping component of this study provided insight into trapping methods and baits useful for salt marshes and coastal forests. Fitch traps were G26 Southeastern Naturalist Vol. 11, No. 1 significantly less effective at capturing small mammals compared to Sherman traps. The mechanical aspects of Fitch traps may be problematic. The opening is slightly smaller than that of the Sherman trap, which might deter animals from entering. The gravity (as opposed to spring-loaded) mechanism will open the door if the trap is turned upside down, which could allow trapped animals to escape. Furthermore, Hayes et al. (1996) suggest that captured individuals may be able to escape by prying the door open. While the more open, mesh appearance may attract certain small-mammal species, it may also be a deterrent for others preferring the more closed, tunnel-like appearance of the Sherman trap. It is unclear if mechanical issues, appearance of the trap to the small mammals or a combination thereof, were the reason for poor performance. We found no significant differences among baits, although raw count data do indicate higher captures using oats followed by scratch feed. There may have been some bias introduced, as baits were not randomly deployed. Traphappy animals discovering a bait pile may have returned repeatedly despite the selection of bait. Oats were not ideal, as they tended to become soggy, penetrating crevices in the trap. This tendency made it difficult to clean the traps, and sometimes affected the spring mechanism of Sherman traps. Suet tended to become rancid and attract non-target animals; its oily nature made it difficult to clean. For these reasons, we preferred dry scratch feed or sunflower seeds. They were water resistant, easy to use, and did not affect the traps negatively in any way. Thus, Sherman traps attached to foam boards for floatation and baited with either scratch feed or sunflower seeds, appears to be the method of choice for direct attempts to capture small mammals in a marsh environment. The use of cameras as an indirect method in upland forests warrants further investigation. Acknowledgments The use of trade, product, industry or firm names, as well as products, software or models, whether commercially available or not, is for informative purposes only and does not constitute an endorsement by the US Government or the US Geological Survey. We would like to acknowledge Billy Brooks (USFWS Jacksonville), LSNWR refuge staff, Simon Fitzwilliam, Althea Hotaling, Rod Hunt, Patrick McElhone, Brandon Miller, John Hayes, (UF Gainesville), Chuck Hunter (FWS Atlanta), and Robert Rose (Old Dominion University). Literature Cited Barret, R.H. 1983. Smoked aluminum track plots for determining furbearer distribution and relative abundance. California Fish and Game 69:188–190. Boonstra, R., M. Kanter, and C.J. Krebs. 1992. A tracking technique to locate small mammals at low densities. Journal of Mammalogy 73(3):683–685. Cain, J.W. K.S. Smallwood, M.L. Morrison, and H.L. Loffland. 2006. Influence of mammal activity on nesting success of passerines. Journal of Wildlife Management 70(2):522–531. Connors, M.J., E.M. Schauber, A. Forbes, C.G. Jones, B.J. Goodwin, and R.S. Ostfeld. 2005. Use of track plates to quantify predation risk at small spatial scales. Journal of Mammalogy 86(5):991–996. 2012 M.A. DeSa, C.L. Zweig, H. F. Percival, W.M. Kitchens, and J.W. Kasbohm G27 Cutler, T.L., and D.E. Swann. Using remote photography in wildlife ecology: A review. Wildlife Society Bulletin 27(3):571–581. DeBondi, N., J.G. White, M. Stevens, and R.Cooke. 2010. A comparison of the effectiveness of camera trapping and live trapping for sampling terrestrial small-mammal communities. Wildlife Research 37(6):456–465. Doonan, T.J., and G.L. Morgan. 2002. Abundance and distribution of the Florida Salt Marsh Vole (Microtus pennsylvanicus dukecampbellli) (Project 9314221 1000). Unpublished Final Project Report. Bureau of Wildlife Diversity Conservation, Florida Fish and Wildlife Conservation Commission. Olustee, FL. Drennan, J.E., P.B. Norris, and L. Dodd. 1998. Use of track stations to index abundance of sciurids. Journal of Mammalogy 79(1):352–359. Eulinger, K.G., and M.S. Burt. 2011. Comparison of captures between Sherman live traps and Museum Special kill traps. Southwestern Naturalist 56(2):241–256. Foran, D.R., S.C. Minta, and K.S. Heinemeyer. 1997. DNA-based analysis of hair to identify species and individuals for population research and monitoring. Wildlife Society Bulletin 25(4):840–847. Gedan, K.B., C.M. Crain, and M.D. Bertness. 2009. Small-mammal herbivore control of secondary succession in New England tidal marshes. Ecology 90(2):430–440. Halfpenny, J. 1986. A Field Guide to Mammal Tracking in North America. Big Earth Publishing, Boulder, CO. 161pp. Hayes, J.P., M.D. Adam, R.G. Anthony, and J.W. Witt. 1996. Comparison of the effectiveness of Sherman and modified Fitch live-traps for capture of small mammals. Northwestern Naturalist 77(2):40–43. Hotaling, A., H.F. Percival, W.M. Kitchens, and J.W. Kasbohm. 2010. The persistence of endangered Florida Salt Marsh Voles in salt marshes of the central Florida Gulf Coast. Southeastern Naturalist 9(4):795–802. Hubbs, A.H., T. Karels, and R. Boonstra. 2000. Indices of population size for burrowing mammals. The Journal of Wildlife Management 64(1):296–301. Justice, K.E. 1961. A new method for measuring home ranges of small mammals. Journal of Mammalogy 42(4):462–470. Karanth, K.U., and J.D. Nichols. 1998. Estimation of tiger densities in India using photographic captures and recaptures. Ecology 79(8):2852–2862. King, C.M., R.M. McDonald, R.D. Martin, G.W. Tempero, and S.J. Holmes. 2007. Longterm automated monitoring of the distribution of small carnivores. Wildlife Research 34:140–148. Kraig, S.E., S.M. Roels, and M.L. Thies. 2010. Effectiveness of chemical repellents in deterring Red Imported Fire Ants (Solenopsis invicta) from Sherman live traps. The Southwestern Naturalist 55(2):203–206. Mabee, T.J. 1998. A weather-resistant tracking tube for small mammals. Wildlife Society Bulletin 26(3):571–574. Masser, M.P., and W.E. Grant. 1986. Fire ant-induced trap mortality of small mammals in east-central Texas. The Southwestern Naturalist 31(4):540–54. McKelvey, K.S., J. Von Kienast, K.B. Aubry, G.M. Koehler, B.T. Maletzke, J.R. Squires, E.L. Lindquist, S. Loch, and M.K. Schwartz. 2006. DNA analysis of hair and scat collected along snow tracks to document the presence of Canada Lynx. Wildlife Society Bulletin 34(2):451–455. Montague, C.L., and R.G. Wiegert. 1990. Salt Marshes. Pp. 481–516, In R.L. Myers and John J. Ewel (Eds.). Ecosystems of Florida. University of Central Florida Press, Gainesville, FL. 765 pp. G28 Southeastern Naturalist Vol. 11, No. 1 Negus, N.C., E. Gould, and R.K. Chipman. 1961. Ecology of the Rice Rat, Oryzomys palustris (Harlan), on Breton Island, Gulf of Mexico, with a critique of the social stress theory. Tulane Studies in Zoology 8(4):95–123. O’Connell, A.F., and L.L. Bailey. 2010. Inference for occupancy and occupancy dynamics. Pp. 191–202, In A.F. O’Connell, J.D. Nichols, and U.K. Karanth (Eds.). Camera Traps in Animal Ecology: Methods and Analyses. Springer, New York, NY. 280 pp. Osterber, D.M. 1962. Activity of small mammals as recorded by a photographic device. Journal of Mammalogy 43(2):219–229. Raabe, E.A., and R.P. Stumpf. 1997. Assessment of acreage and vegetation change in Florida’s Big Bend tidal wetlands using satellite imagery: Proceedings of the fourth international conference on remote sensing for marine and coastal environments, Orlando, FL, 17–19 March, 1997, I:84–93. . Reed, J.Z., D.J. Tollit, P.M. Thompson, and W. Amos. 1997. Molecular scatology: The use of molecular genetic analysis to assign species, sex, and individual identity to seal faeces. Molecular Ecology 6(3): 225–234. Rose, R.K. 1994. Instructions for building two live traps for small mammals. Virginia Journal of Science 45(3):151–158. Slade, N.A., M.A. Eifler, N.M. Gruenhagen, and A.L. Davelos. 1993. Differential effectiveness of standard and long Sherman livetraps in capturing small mammals. Journal of Mammalogy 74:156–161. Srbek-Araujo, A.C., and A.G. Chiarello. 2005. Is camera-trapping an efficient method for surveying mammals in Neotropical forests? A case study in southeastern Brazil. Journal of Tropical Ecology 21:121–125. Stout, J.P. 1984. The ecology of irregularly flooded salt marshes of the northeastern Gulf of Mexico: A community profile. US Fish and Wildlife Service, Biological Report 85(7.1). 98 pp. Available online at http://www.nwrc.usgs.gov/diglib.htm#comprofs. Accessed 15 February 2011. Trolle, M., and M. Kéry. 2003. Estimation of Ocelot density in the Pantanal using capturerecapture analysis of camera trapping data. Journal of Mammalogy 84(2):607–614. Vine, S.J., M.S. Crowther, S.J. Lapidge, C.R. Dickman, N. Mooney, M.P. Piggott, and A.W. English. 2009. Comparison of methods to detect rare and cryptic species: A case study using the Red Fox (Vulpes vulples). Wildlife Research 26:436–446. Wiener, J.G., and M.H. Smith. 1972. Relative efficiencies of four small-mammal traps. Journal of Mammalogy 53:868–873. Williams, D.F., and S.E. Braun. 1983. Comparison of pitfall and conventional traps for sampling small-mammal populations. The Journal of Wildlife Management 47(3):841–845. Wolfe, J.L. 1982. Oryzomys palustris. Mammalian Species 176:1–5. Wolfe, J.L. 1985. Population ecology of the Rice Rat (Oryzomys palustris) in a coastal marsh. Journal of Zoology (London) 205:235–244. Woods, C.A. 1992. Endangered: Florida Saltmarsh Vole (Microtus pennsylvanicus dukecampbelli). Pp. 131–139, In S.R. Humphrey (Ed.). Rare and Endangered Biota of Florida, Vol. 1: Mammals. University Press of Florida. Gainesville, FL. 392 pp. Woods, C.A., W. Post, and C.W. Kilpatrick. 1982. Microtus pennsylvanicus (Rodentia: Muridae) in Florida: A Pleistocene relict in a coastal saltmarsh. Bulletin of the Florida State Museum Biological Sciences 28(2):25–52.