SEARCH

SEARCH BY CITATION

Keywords:

  • aversive conditioning;
  • behavioural modification;
  • feral cat;
  • island fox;
  • live-trapping;
  • nontarget recaptures;
  • predator management;
  • reward removal;
  • San Clemente Island;
  • Urocyon littoralis

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

1. Live-trapping is a fundamental tool in the study of wildlife species and populations. Capture of nontarget species is an inherent side-effect of trapping animals that is inefficient and potentially detrimental to individuals and populations. Reducing recapture of nontarget species will increase the efficiency of long-term studies and projects, and minimize unwanted impacts.

2. During the initial stages of a 3-year feral cat Felis catus control programme on San Clemente Island, California, USA, we experienced high recapture rates of nontarget endangered San Clemente Island foxes Urocyon littoralis clementae, impacting project efficiency and potentially foxes.

3. We investigated whether we could modify our baiting strategy to reduce fox recaptures using behavioural modification techniques, while simultaneously maintaining the ability to capture feral cats. We tested two strategies, aversive conditioning and reward removal, to reduce fox recapture rates.

4. Using lithium chloride (LC) as an aversive agent, we reduced fox recapture rates more than 10 times compared with rates using control bait. Using the reward removal technique, we reduced fox recapture rates almost 20 times compared with control baits. Neither behavioural modification technique deterred cats from entering traps, and our results suggest that by reducing fox recaptures, we improved our ability to capture cats.

5.Synthesis and applications. These novel baiting strategies have wide applicability in wildlife research and management. The efficiency of long-term monitoring or control programmes may be enhanced by discouraging recapture of ‘trap happy’ animals. Potential impacts to threatened species confined in traps could be minimized by reducing recaptures. Sampling of rare species in biodiversity studies could be improved by reducing recapture of competitively superior or numerically dominant species.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

To study and manage most species of mammals, it is essential to capture individuals. It is important to maximize captures of target species and minimize captures of nontarget species for reasons of economy and to develop robust datasets. Box traps are often preferred over other types (e.g. foothold) because of a reduced risk of injury and the greater protection box traps afford from weather and predators. Although box traps have been shown to be less traumatic to captured animals than foothold traps or snares (Kreeger et al. 1990; Mowat, Slough & Rivard 1994), they are not a benign method (Proulx & Barrett 1989; Putman 1995). Confinement in box traps can cause physiological stress in many species of mammals (White et al. 1991; Fletcher & Boonstra 2006; Schütz et al. 2006), often resulting in physical injury (e.g. broken jaws, teeth or claws) (Neighbor et al. 1991; Covell 1992; Mowat, Slough & Rivard 1994; R. B. Phillips, unpublished data). Such traumas often are dismissed as unimportant side-effects of livetrapping (Mowat, Slough & Rivard 1994); however, trap-induced stress is considered a major cause of trap mortality for rodents (Perrin 1975; Montgomery 1980; Gurnell 1982).

Recapture of nontarget species is potentially detrimental to individuals and populations, and it adversely affects captures of target species, reducing efficiency (Reagan et al. 2002; Way et al. 2002). Reducing nontarget recaptures will increase trapping efficiency for many projects that focus on a particular species or suite of species: predator management (Witmer et al. 1996; Neuman et al. 2004), rabies monitoring and control (Rosatte et al. 1990; Krebs, Wheeling & Childs 2003), and insular invasive species eradication (Nogales et al. 2004; Howald et al. 2007). These projects may span several years and are active continuously for several months, often during the breeding seasons of nontarget species.

While conducting a feral cat Felis catus Linneaus management project for the protection of the federally endangered San Clemente loggerhead shrike Lanius ludovicianus mearnsii Miller, we captured nontarget island foxes Urocyon littoralis clementae Merriam. This subspecies is listed as threatened by the state of California (California Department of Fish and Game 1987) and has recently experienced a severe population decline (Roemer & Wayne 2003). Island foxes have high capture (0·24–0·60) and recapture (0·31–0·87) probabilities and often become ‘trap happy’ (Roemer et al. 1994; R. B. Phillips unpublished data). Because of the long-term nature of our project and the critical status of the island fox, it was important to reduce the potential impact to the population from frequent recapture. We hypothesized that by creating a negative or neutral trap experience using aversive conditioning or ‘reward removal’, respectively, we could reduce recaptures of nontarget island foxes. Aversive conditioning is a technique that applies a negative stimulus (e.g. pain, irritation) to an animal engaged in an unwanted behaviour (Brush 1971). If successful, the animal associates the negative stimulus with the unwanted behaviour and ceases that behaviour. Attempts to alleviate a variety of wildlife damage problems using aversive conditioning have yielded equivocal results (Conover 1997). It has been used successfully to reduce egg predation (Massei, Lyon & Cowan 2002) and human-bear conflicts (Mazur 2010), and to prevent an endangered predator from foraging on toxic prey (O’Donnell, Webb & Shine 2010). Reward removal is a simple concept, whereby one attempts to stop an undesirable behaviour by removing the stimulus that prompts that behaviour. This method also has been effective in alleviating wildlife–human conflicts (Dalle-Molle & Van Horn 1989; Manley & Williams 1998).

The practice of wildlife management is inherently intrusive; however, biologists have the responsibility to minimize the impact of their studies on nontarget animals. In this paper, we investigate whether aversive conditioning and reward removal techniques can be used to reduce recapture rates of nontarget species without negatively affecting management of the target species. We then evaluate the different response of foxes to each baiting strategy to assess the potential application of the technique for reducing recaptures of island foxes and of nontarget species in general.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This work was conducted from 1992 to 1994. After 1994, use of methods to reduce fox recaptures was discontinued because of staff turnover and a change in project administration. A thorough search of the conventional and grey literature yielded no studies where aversive conditioning or reward removal have been used to reduce recaptures of nontarget species.

Trapping and Treatment Application

We conducted field trials on San Clemente Island, California, USA (32°55′N, 118°30′W), a 146-km2 island located 102 km west-northwest of San Diego, California. The island is a military reserve, with the southern third used weekly for naval activities. Work was conducted on the southern half of the island in the predominant habitats of grassland, maritime desert scrub and woodland (Raven 1963; Philbrick & Haller 1977).

In all 3 years, fieldwork started around 1 February, ending in July 1992, August 1993 and June 1994, to coincide with the shrike breeding season. Our efforts were focused on protecting shrikes from feral cats during breeding, thus, we began trapping in an area when the shrike monitoring team verified nesting activity. Similarly, we stopped trapping in an area when shrike fledglings were independent or a nest failed.

We used 23 × 23 × 66 cm Tomahawk™ cage traps. Our standard bait was a 1:1:1 mixture of canned cat food (beef and liver flavour), canned tuna and raw hamburger, applied in 50 g quantities at the rear of the trap. We placed traps along two roughly parallel transects or in a single circular perimeter around the shrike nest, depending upon the topography of the area. Traps were placed a minimum of 200 m from the shrike nest and at about 250 m intervals. The number of traps placed for all areas ranged from 8 to 20.

We first anaesthetized captured cats by an intramuscular injection containing a mixture of ketamine hydrochloride (Ketaset™, 20 mg kg−1 body weight) and xylazine hydrochloride (Rompun™, 1 mg kg−1), then euthanized by intracardiac injection of sodium pentabarbitol (Terminol-II C-II™, 1 mL/4·5 kg); we handled foxes without anaesthesia. Each fox was marked with an ear tag containing a unique alphanumeric code.

In 1992, our trials were exploratory. We evaluated the efficacy of the aversive agent, lithium chloride (LC), an odourless salt (Lithium chloride 2008), as a bait additive to reduce fox recaptures. Relative to other chemicals, such as carbachol, which is often fatal, LC is mild and produces short-term effects (Conover 1997). In coyotes Canis latrans Say, concentrations of LC sufficient to cause severe nausea, but not emesis, produced the strongest conditioned aversion (Burns & Connolly 1980; Burns 1983). LC had not been previously used as an aversive agent on a canid comparable in size to the island fox; therefore, we tested various concentrations of LC in the bait mixture. We used dosage levels presented in Burns (1983) as a baseline protocol. Initially, we mixed 20 mL of a 10% LC solution (10 g LC/100 mL tap water) with 500 g of the standard bait. This concentration yielded 0·1 g of LC/50 g bait portion/trap; a dosage approximately equal to 100 mg of LC per kg of body weight (males = 2·0 kg, females = 1·8 kg) (Roemer et al. 2004). We applied bait mixed with 3 LC concentrations (100, 150 mg and 200 mg kg−1), sequentially increasing the dosage if no reduction in fox recapture rates was detected. The number of days a specific dosage was applied in an area varied from two to 11. To test the efficacy of reward removal on reducing fox recaptures, we perforated the tops of 170 g cat food canisters with 4 mm holes. We placed a single canister in the rear floor of each trap and replaced canisters every 3–4 days. Thus, foxes and cats were enticed into entering traps, but were prevented from consuming bait. The standard bait mixture served as the control in the field trials.

In 1992, we applied control bait, perforated canisters (PC) and LC in three, two and two areas, respectively (Table 1). Because of a delay in receiving LC, we were unable to test reward removal and aversive conditioning methods concurrently. Application of control baits and PC began in early February and LC in mid-April.

Table 1.   Methods used to reduce nontarget captures of island foxes Urocyon littoralis clementae on San Clemente Island, California, USA, between 1992 and 1994 during a feral cat Felis catus management programme
YearMethodNumber of areas trappedTrapnightsFox captures*Fox recaptures†Cat captures
  1. Recapture rates for foxes and capture rates for cats were evaluated using different baiting strategies (see Materials and methods): Ctl, control bait; LC, lithium chloride; PC, perforated canisters.

  2. *Fox captures represent individual foxes capture under different baiting strategies.

  3. †Fox recaptures do not include new captures and first capture of an individual under a new baiting strategy.

1992Ctl 35427130023
PC 2573243518
LC-100 214016260
LC-150  15714171
LC-200  1901371
1993Ctl 728815967652
LC-Ctl(pre-switch)21231273245
(post-switch)13772519210
1994Ctl-PC(pre-switch)310072421113
(post-switch)43619331
PC-Ctl(pre-switch)287671112
(post-switch)449171241

In 1993, we used LC treated bait applied at the 200 mg dosage (based on results from 1992) and the control bait was as in 1992. We trapped in nine different areas (Table 1). In seven areas, we used only control bait. In two other areas, we used a crossover design where the LC was applied first (‘pre-switch’), followed by control bait (‘post-switch’). We alternated the LC baits with control baits to evaluate the response time in foxes to a change in baiting strategy.

In 1994, we used the reward removal strategy, refining the 1992 method by using 5 × 5 cm, 100 mL plastic canisters instead of cat food canisters. We drilled the lids and the upper perimeter of the canisters with 4-mm holes. We then permanently wired a lid, thread side down, to the rear ceiling of each trap. To provide a visual and additional olfactory cue, we also placed a fish oil scented cotton ball in the bottom rear of each trap. Canisters were filled with control bait, which was the same formulation used in previous years. We trapped in five areas, applying both PC and control baits in a crossover design (Table 1).

Statistical Analyses

We evaluated the three strategies by comparing the cat capture rate and fox recapture rate (omitting new fox captures) using each bait. We defined new captures as the first capture for an individual under a specific baiting strategy. We expressed rate as the number of foxes recaptured or cats captured per 100 trapnights (corrected for nontarget captures and sprung traps) (Nelson & Clark 1973; Beauvais & Buskirk 1999). In 1992, we calculated the overall mean fox recapture and cat capture rate for each treatment and area combination. We then compared these rates between the LC, for the three dosages, the PC and the control bait using descriptive statistics (Zar 1984; Cherry 1998).

For 1993, we calculated mean daily fox recapture and cat capture rates. We used a general linear mixed model to compare fox recapture rates using control bait (nonswitch areas; Table 1) with LC bait. We assigned treatment (control or LC) as a categorical explanatory variable and day as a continuous explanatory variable. The interaction of these two variables was included in the model. We defined sites as a random-effects blocking factor. To assess fox response time to a change in baiting strategy, we then compared fox recapture rates using LC bait in the pre-switch period with rates using control bait in the post-switch period (Table 1) using a piecewise linear regression. The breakpoint was set at the bait switch date (between days 41 and 42).

We compared 1994 fox recapture rates with control baits (pre-switch periods) to PC (pre-switch periods; Table 1) using a general linear mixed model parameterized as for 1993 fox data. We then examined bait application sequence (control to PC and PC to control) using a general linear mixed model with bait and sequence as categorical explanatory variables and day as a continuous explanatory variable. Site was defined as a random-effects blocking factor. The statistical model simultaneously fit a separate piecewise linear regression for each order. The breakpoint was set at the bait switch date (between days 43 and 44). Separate residual variances were estimated for the four combinations of treatment and order (i.e. variances were heterogeneous).

To analyse the 1993 and 1994 cat capture rates between treatment and control baits, we used only the first 3 weeks of data because cat capture rates declined to nearly zero after the first 2–3 weeks of trapping (Phillips & Schmidt 1996; Harding, Doak & Albertson 2001). We used a general linear mixed model with treatment as a categorical explanatory variable and day as a continuous explanatory variable. The interaction of these two variables was included initially in the model, but was not significant and subsequently was omitted. This forces the two regression lines to have the same slope.

Finally, we examined the response of individual foxes to the three baiting strategies in 1993 and 1994. We calculated the number of times each fox was recaptured using a particular bait each year. For areas where we switched bait, we examined the response of individual foxes to the order in which baiting strategies were applied.

For all GLM and the piecewise linear regressions, the response variable was square-root transformed prior to analysis to normalize the data. All computations were carried out using PROC MIXED in SAS Release 7.0 (SAS Institute 1996). For capture and recapture rates, we present the mean and standard error (SE), and for number of individuals and times captured, we present the mean, standard deviation (SD) and sample size (n).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Mean fox recapture rates in 1992 were highest using control baits (Fig. 1, Table 1). The LC-200 dosage appeared to be most effective at reducing fox recapture rates, with rates decreasing substantially in comparison with LC-150. Recapture rates were greatly reduced using PC compared with control bait and only slightly lower than LC-200. Mean cat capture rates were higher using PC than LC-200 (Fig. 1), but capture rates using the LC-200 were similar to those using the control bait.

image

Figure 1.  Fox recapture and cat capture rates in 1992 using three baiting strategies, lithium chloride (LC; three dosage levels), perforated canisters (PC) and control bait (Ctl). Data are means +1 SE.

Download figure to PowerPoint

In 1993, overall fox recapture rates were more than 10 times lower using LC (inline image = 2·75, SE = 0·44, n = 78) than with the control bait in nonswitch areas (Table 1, inline image = 30·22, SE = 1·85, n = 204) (F1, 7 = 9·27, = 0·019). In areas where only the control bait was used, fox recapture rates increased over time (F40, 178 = 2·15, < 0·001, Fig. 2), whereas using LC, they remained relatively static. In the bait switch areas, fox recapture rates increased following the change from LC to the control bait (t = 6·89, d.f. = 156, < 0·001, Fig. 3). However, there was about a 10-day lag time before the foxes responded to the change in bait. As in 1992, LC did not limit our ability to capture cats. For the first 3 weeks of trapping, cat capture rates were two times as high using LC (inline image = 6·82, SE = 1·23, = 42) than control baits (inline image = 3·58, SE = 0·62, n = 128) (F1, 7 = 7·26, = 0·039).

image

Figure 2.  Fox recapture rates in 1993 in areas using aversive conditioning (lithium chloride) and control (Ctl) baiting strategies. Error bars are ±1 SE.

Download figure to PowerPoint

image

Figure 3.  Fox recapture rates in 1993 in areas using aversive conditioning baiting strategy [lithium chloride (LC)] followed by control bait (Ctl). Bait was switched from LC to Ctl on day 41 (denoted by arrow).

Download figure to PowerPoint

The response of individual foxes to LC corroborates the results of our overall capture rates. In areas where only control bait was used, individual foxes were recaptured frequently (inline image = 11·46, SD = 14·34, = 59). In the two areas where we applied LC followed by control bait, foxes were recaptured less often during the period when LC was applied (inline image = 1·12, SD = 1·67, = 27) compared with the period when control bait was applied (inline image = 7·68, SD = 7·93, = 25). Of the 27 foxes captured using LC in the pre-switch period, 16 were captured in the post-switch period using the control bait.

In 1994, during the pre-switch periods (Table 1), overall fox recapture rates were almost 20 times higher using the control bait (inline image = 26·84, SE = 2·27, = 104) than the PC (inline image = 1·38, SE = 0·43, n = 63) (F1, 3 = 157·97, < 0·001). During the pre-switch period in the control to PC areas, mean fox recapture rates tended to increase over time, but declined following application of the treatment bait (F1, 234 = 34·01, < 0·001, Fig. 4). In contrast, mean fox recapture rates during the pre-switch period in the PC to control areas remained static and near zero during use of treatment bait, but increased almost immediately following the switch to the control bait (F1, 234 = 116·34, < 0·001). The trends in fox recapture rates during the pre-switch period differed between control and treatment baits (Fig. 4; F1, 234 = 11·05, < 0·001), as did recapture rates during the post-switch phase (F1, 234 = 148·94, < 0·001). Cat capture rates during the first 3 weeks were similar (F1, 3 = 0·12, = 0·729) for areas where PC (inline image = 1·75, SE = 0·65, n = 42) and control bait (inline image = 2·29, SE = 0·68, = 63) were used.

image

Figure 4.  Fox recapture rates in 1994 in areas using reward removal (perforated canisters; PC) and control bait (Ctl). Baits were switched on day 43 (denoted by arrow). Solid line indicates areas where PC preceded Ctl, and dashed line indicates areas where Ctl preceded PC.

Download figure to PowerPoint

In the control to PC areas, individual foxes were recaptured more frequently in the pre-switch period (inline image = 8·42, SD = 9·53, = 24) than in the post-switch period (inline image = 1·42, SD = 2·36, = 19). In contrast, in the PC to control areas, foxes were recaptured less often in the pre-switch period (inline image = 1·57, SD = 1·13, = 7) than in the post-switch period (inline image = 4·88, SD = 5·33, = 17). Of the 24 foxes captured in the control to PC areas, pre-switch period, 17 were captured in the post-switch period. Of the seven foxes captured in the PC to control areas, pre-switch period, all were captured in the post-switch period.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We demonstrate that recapture rates of a nontarget species can be greatly reduced using either aversive conditioning or reward removal techniques. In our preliminary trials in 1992, we observed a corresponding drop in fox recapture rates with each subsequent increase in the LC dosage. Using either aversive conditioning or reward removal, we were able to maintain fox recapture rates between 0% and 10% over several months. In contrast, using control bait over the same period, recapture rates tended to increase over time from 0% to 80% in some areas.

Although both aversive conditioning and reward removal were highly effective in reducing overall fox recapture rates, the two methods elicited different responses in foxes. The use of LC created an aversion to recapture that persisted for at least 10 days as exhibited by the delayed increase in recapture rates following the switch from LC to control bait. In some individual foxes, the aversion appeared to persist beyond the duration of our trial. Eleven of the 27 foxes captured using LC in the pre-switch phase were not recaptured following the switch to control bait. The strength of the aversion may be because of the inability of foxes to distinguish between treatment and control baits from outside the trap and the severity of the nausea after consuming the bait (Conover 1989). In contrast, when we switched from PC to control bait, the foxes responded almost immediately, with a rapid increase in fox recapture rate, and all (7 of 7) of the individual foxes captured in the pre-switch phase were captured in the post-switch phase. Foxes probably relied on a combination of olfactory and visual cues to distinguish between baiting strategies. The odour of the PC and control strategies was similar, as the bait for each was the same. The control, however, lacked the fish oil scented cotton ball, both a visual and olfactory cue. Overall, the two strategies were visually distinct, a white plastic vial suspended in the corner compared with a dollop of meat on the floor of the trap. Regardless of the mechanism, foxes appeared able to distinguish strategies.

Unfortunately, the noticeable differences in the two strategies appear to have not only affected the response times of foxes to a change in baiting strategy but may also have affected initial captures in the pre-switch phase. When we applied PC prior to control bait, we captured only seven of the 17 previously marked individuals; whereas when control bait was applied prior to PC, we captured 24 of the 26 marked foxes. Similarly, using LC followed by control bait, we captured 29 of the 36 marked foxes.

Despite some foxes appearing to avoid capture based on the presence of PC in 1994, this method did not appear to affect our overall ability to capture cats. In 1992, cat capture rates were highest using PC, and in 1994, there was no difference between cat capture rates using PC and control baits. Equally important, in the post-switch period, we observed no increase in cat captures when we switched from PC to control baits, suggesting cats were not avoiding capture based on the presence of PC.

While reward removal method was effective for capturing cats, so too was LC. In 1992, cat capture rates using LC were similar to those using control baits. Although cat capture rates were higher using PC than with LC, this was likely to be an artefact of the timing of bait application. We applied PC at the beginning of the season, whereas we began using LC after more than 2 months of feral cat removal. In predator management projects, the majority of target animals are captured in the initial period (Phillips & Schmidt 1996; Harding, Doak & Albertson 2001). Notably, cat capture rates using PC were higher than those using control bait, which were applied concurrently. This suggests that reducing the number of traps occupied by foxes increased their availability to cats. We observed similar results in 1993, with cat capture rates twice as high using LC vs. control baits. The improvement we observed in cat capture rates in 1993 between treatment and control baits compared with no difference in 1994 is probably due to an area-wide suppression in the cat population from 3 years of feral cat control (Phillips & Schmidt 1996).

Our success in creating a conditioned aversion in our subjects (foxes) compared with other studies is likely to be due to multiple factors. First, in contrast to other studies, we were not attempting to counter a strong instinctual behaviour, such as predation (Burns & Connolly 1980). Secondly, in most research on conditioned aversion, the objective has been to create an aversion to the model (e.g. killing live sheep) based on a negative experience with a mimic (e.g. consuming mutton containing an aversive agent) (Conover 1997). A conditioned aversion is created only if the animal easily associates the mimic to the model. In our study, we were attempting to create an aversion directly to the model, a connection readily generalized. Finally, a conditioned aversion is more easily generalized if the chemical is difficult to detect (Conover 1989). Because LC is odourless and the pre- and post-switch baits were visually indistinguishable, foxes were unable to determine the status of the bait without being captured.

Although we made no attempt to document it, reducing fox recapture rates increased our project efficiency. Field personnel spent fewer hours resetting traps and handling foxes, thus allowing us to monitor a greater number of traps, and with fewer traps occupied by foxes more were available to cats. More importantly, reducing recaptures probably benefitted individual foxes and the fox population. As with many management projects, our field activities were conducted continuously over several months (5–7). Given the duration of the project and the high recapture rates of foxes (Roemer et al. 1994), there was great potential to negatively impact foxes. By having fewer recaptures, we minimized the incidence of trap-related trauma, such as broken teeth and elevated stress levels (White et al. 1991; Mowat, Slough & Rivard 1994; Fletcher & Boonstra 2006; Schütz et al. 2006). Foxes also spent less time confined in traps thereby helping to maintain social systems and since our trapping spanned the fox breeding season, it is likely that we minimized the confinement of females with dependent pups.

The continued application of our methods on San Clemente Island following this study may have led to more effective control of feral cats, potentially resulting in reduced predation on juvenile shrikes. Resources saved because of a more efficient cat control programme could have been redirected toward enhanced management practices (e.g. shock collars for foxes, native shrub establishment) aimed at preventing fox predation on juvenile shrikes (Cooper et al. 2001; Roemer & Wayne 2003). An increase in the shrike population because of greater survival of juvenile shrikes also may have permitted managers to accept a low, but natural level of fox predation on shrikes, obviating the need to remove foxes from shrike nesting areas.

Our recapture reduction methods should be transferable to other species, thus allowing managers to reduce recaptures of nontarget species without affecting their ability to capture the target species. Each method has distinct properties that provides adaptability in its application. Reward removal has the advantage of being able to be activated or deactivated rapidly, but some individuals may avoid capture. Conversely, the strong conditioned aversion created by using LC results in a longer latent period when foxes appear unable to detect the status of the trap. Our recapture reduction strategies have wide-scale potential applicability in wildlife studies. Wildlife managers working on long-term monitoring or control projects (e.g. eradication and control of invasive species on islands) could apply our strategies to achieve an increase in efficiency. Biologists involved with projects related to species of concern may be able to use these methods to reduce recaptures of nontarget species. Similarly, researchers conducting biodiversity studies could use our strategies to reduce the recapture rate of competitively superior or numerically dominant species, thus allowing subordinate or rare species to be sampled. For San Clemente Island and the other California Channel Islands, we recommend our recapture reduction methods be employed in any long-term feral cat management programme. With the current precarious status of foxes on most islands (Roemer et al. 2004), the potential impacts to nontarget foxes from frequent recapture should be minimized. However, if our methods are employed, we recommend the fox population be monitored concurrently. In the event some foxes prove resistant to recapture reduction it is imperative to be able to assess the situation.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank Melissa Booker, Charlotte Coulter, Steve Fairbairn, Laura Felicetti, Frans Joula, Timothy Leifer, Chris Lee, Phil Moggridge, Bruce Rodrigues, Adam Rosenberg, Craig Thompson, Adam Walker and Christine Wilder for assistance in the field. The U. S. Navy, North Island Naval Air Station, Natural Resources Office staff provided logistical support. We thank Richard J. Burns, Michael R. Conover and Carl R. Gustavson for their input on aversive conditioning theory and techniques and Robert H. Schmidt for project advice. Susan Durham provided assistance with statistical analysis. Two reviewers, Al Glen and Gary Roemer, provided helpful comments and suggestions that improved the manuscript. This research was approved and conducted under the Utah State University Institutional Animal Care and Use Committee application # 574. The findings and conclusions in the article are those of the author(s) and do not necessarily represent the views of the US Fish and Wildlife Service.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Beauvais, G.P. & Buskirk, S.W. (1999) Modifying estimates of sampling effort to account for sprung traps. Wildlife Society Bulletin, 27, 3943.
  • Brush, F. (1971) Aversive Conditioning and Learning. Academic Press, New York, NY.
  • Burns, R.J. (1983) Microencapsulated lithium chloride bait aversion did not stop coyote predation on sheep. Journal of Wildlife Management, 47, 10101017.
  • Burns, R.J. & Connolly, G.E. (1980) Lithium chloride bait aversion did not influence prey-killing by coyotes. Proceedings of the Vertebrate Pest Conference, 9, 200204.
  • California Department of Fish and Game. (1987) Five-year Status Report on the Island Fox (Urocyon littoralis). California Department of Fish and Game, Sacramento, CA.
  • Cherry, S. (1998) Statistical tests in publications of The Wildlife Society. Wildlife Society Bulletin, 26, 947953.
  • Conover, M.R. (1989) Potential compounds for establishing conditioned food aversions in raccoons. Wildlife Society Bulletin, 17, 430435.
  • Conover, M.R. (1997) Behavioral principles governing conditioned food aversions based on deceptions. Repellents in Wildlife Management (ed. J.R. Mason), pp. 2941. Proceedings of the Second Denver Wildlife Research Center Special Symposium, Denver, CO.
  • Cooper, D., Kershmer, M.E.L., Schmidt, G.A. & Garcelon, D.K. (2001) San Clemente Loggerhead Shrike Predator Research and Management Program –2000. Institute for Wildlife Studies, Arcata, California. Final report. US Navy, Natural Resources Management Branch, San Diego, CA.
  • Covell, D.F. (1992) Ecology of the swift fox (Vulpes velox) in southeastern Colorado. M.S. thesis, University of Wisconsin, Madison, WI.
  • Dalle-Molle, J.L. & Van Horn, J.C. (1989) Bear-people conflict management in Denali National Park, Alaska. Bear-people Conflicts: Proceedings of a Symposium on Management Strategies (ed. M. Bromley), pp. 121127. Department of Renewable Resources, Yellowknife, Northwest Territories, Canada.
  • Fletcher, Q.E. & Boonstra, R. (2006) Impact of live trapping on the stress response of the meadow vole (Microtus pennsylvanicus). Journal of Zoology, 270, 473478.
  • Gurnell, J. (1982) Trap deaths in woodland rodents. Acta Theriologica, 27, 139147.
  • Harding, E.K., Doak, D.F. & Albertson, J.D. (2001) Evaluating the effectiveness of predator control: the non-native red fox as a case study. Conservation Biology, 15, 11141122.
  • Howald, G., Donlan, C.J., Galván, J.P., Russell, J.C., Parkes, J., Samaniego, A., Wang, Y., Veitch, D., Genovesi, P., Pascal, M., Saunders, A. & Tershy, B. (2007) Invasive rodent eradication on islands. Conservation Biology, 21, 12581268.
  • Krebs, J.W., Wheeling, J.T. & Childs, J.E. (2003) Public veterinary medicine: public health – Rabies surveillance in the United States during 2002. Journal of the American Veterinary Medical Association, 223, 17361748.
  • Kreeger, T.J., White, P.J., Seal, U.S. & Tester, J.R. (1990) Pathological responses of red foxes to foothold traps. Journal of Wildlife Management, 54, 147160.
  • Lithium chloride (2008) MSDS No. L6697, Mallinckrodt Baker, Inc., Phillipsburg, NJ Available at: http://www.avantormaterials.com/documents/MSDS/USA/English/L6697_msds_us_Default.pdf(accessed on 21 November 2008).
  • Manley, T. & Williams, J. (1998) Bearproofing solid waste containers for grizzly and black bears in Lake and Cascade counties, Montana. Intermountain Journal of Sciences, 4, 101.
  • Massei, G., Lyon, A.J. & Cowan, D.P. (2002) Conditioned taste aversion can reduce egg predation by rats. Journal of Wildlife Management, 66, 11341140.
  • Mazur, R.L. (2010) Does aversive conditioning reduce human-bear conflict? Journal of Wildlife Management, 74, 4854.
  • Montgomery, W.I. (1980) Mortality of small rodents captured in live-traps. Acta Theriologica, 25, 277294.
  • Mowat, G., Slough, B.G. & Rivard, R. (1994) A comparison of three live capturing devices for lynx: capture efficiency and injuries. Wildlife Society Bulletin, 22, 644650.
  • Neighbor, D.S., Ruth, T.K., Skiles Jr, J.R. & McKinney Jr, B.P. (1991) Live-trapping mountain lion. Mountain Lion-Human Interaction: Symposium and Workshop (ed. C.E. Braun), pp. 25. Colorado Division of Wildlife, Denver, CO.
  • Nelson Jr, L. & Clark, F.W. (1973) Correction for sprung traps in catch-effort calculations of trapping results. Journal of Mammalogy, 54, 295298.
  • Neuman, K.K., Page, G.W., Stenzel, L.E., Warriner, J.C. & Warriner, J.S. (2004) Effect of mammalian predator management on snowy plover breeding success. Waterbirds, 27, 257263.
  • Nogales, M., Martín, A., Tershy, B.R., Donlan, C.J., Veitch, D., Puerta, N., Wood, B. & Alonso, J. (2004) A review of feral cat eradications on islands. Conservation Biology, 18, 310319.
  • O’Donnell, S., Webb, J.K. & Shine, R. (2010) Conditioned taste aversion enhances the survival of an endangered predator imperiled by a toxic invader. Journal of Applied Ecology, 47, 558565.
  • Perrin, M.R. (1975) Trap deaths. Acta Theriologica, 20, 167174.
  • Philbrick, R.N. & Haller, J.R. (1977) The Southern California Islands. Santa Barbara Botanic Garden, Santa Barbara, CA.
  • Phillips, R.B. & Schmidt, R.H. (1996) San Clemente Island, California Feral Cat Management Plan. Final report, Utah State University, Logan, UT.
  • Proulx, G. & Barrett, M.W. (1989) Animal welfare concerns and wildlife trapping: ethics, standards and commitments. Transactions of the Western Section of the Wildlife Society, 25, 16.
  • Putman, R.J. (1995) Ethical considerations and animal welfare in ecological field studies. Biodiversity and Conservation, 4, 903915.
  • Raven, P.H. (1963) A flora of San Clemente Island, California. Aliso, 5, 289387.
  • Reagan, S.R., Ertel, J.M., Stinson, P., Yakupzack, P. & Anderson, D. (2002) A passively triggered foot snare design for American black bears to reduce disturbance by non-target animals. Ursus, 13, 317320.
  • Roemer, G.W. & Wayne, R.K. (2003) Conservation in conflict: the tale of two endangered species. Conservation Biology, 17, 12511260.
  • Roemer, G.W., Garcelon, D.K., Coonan, T.J. & Schwemm, C. (1994) The use of capture-recapture methods for estimating, monitoring, and conserving island fox populations. The Fourth California Islands Symposium: Update on Status Resources (eds W.L. Halvorsen & G.J. Maender), pp. 387400. Santa Barbara Museum of Natural History, Santa Barbara, CA.
  • Roemer, G.W., Coonan, T.J., Munson, L. & Wayne, R.K. (2004) Island fox, Urocyon littoralis. Canid action plan for the island fox. Canids: Foxes, Wolves, Jackals, and Dogs. Status Survey and Conservation Action Plan (eds C. Sillero-Zuiri, M. Hoffman & D.W. Macdonald), pp. 97105. International Union for Conservation of Nature/Species Survival Commission, Canid Specialist Group, IUCN Publications Services Unit, Cambridge.
  • Rosatte, R.C., Power, M.J., MacInnes, C.D. & Lawson, K.F. (1990) Rabies control for urban foxes, skunks, and raccoons. Proceedings of the Vertebrate Pest Conference, 14, 160167.
  • SAS Institute. (1996) SAS/STAT User’s Guide. Release 6.12. Volume 1, 4th edn. SAS Institute, Cary, IN.
  • Schütz, K.E., Ågren, E., Amundin, M., Röken, B., Palme, R. & Mörner, T. (2006) Behavioral and physiological responses of trap-induced stress in European badgers. Journal of Wildlife Management, 70, 884891.
  • Way, J.G., Ortega, I.M., Auger, P.J. & Strauss, E.G. (2002) Box-trapping eastern coyotes in southeastern Massachusetts. Wildlife Society Bulletin, 30, 695702.
  • White, P.J., Kreeger, T.J., Seal, U.S. & Tester, J.R. (1991) Pathological responses of red foxes to capture in box traps. Journal of Wildlife Management, 55, 7580.
  • Witmer, G.W., Bucknall, J.L., Fritts, T.H. & Moreno, D.G. (1996) Predator management to protect endangered avian species. Transactions of the North American Wildlife and Natural Resources Conference, 61, 102108.
  • Zar, J.H. (1984) Biostatistical Analysis, 2nd edn. Prentice Hall, Englewood Cliffs, NJ.