Context. Scat genotyping has not been routinely used to measure fox (Vulpes vulpes) abundance and our study sought to provide a benchmark for further technique development and assessment of field methods.Aims. This study sought to provide a comparative assessment of some common methods used to determine fox density and contrast their success with scat DNA genotyping.Methods. DNA recovered from fox scats was used to genotype individual red foxes and determine their abundance at four transects. Population indices were also developed from bait take, scat counts and sand plot tracks using index-manipulationindex (IMI) procedures on the same transects. Known samples of foxes were taken from two treatment transects using cyanide delivered in the M-44 ejector to manipulate the population and to recover foxes at the end of the trial.Key results. Replicated counts on a 41-km-spotlight transect at the field site before and after the population manipulation had low variance and good correlation (r 2 = 0.79, P < 0.01). Scat genotypes revealed 54 foxes in eight days and, when combined with biopsy DNA from recovered foxes, a minimum known to be alive (KTBA) density of between 1.6 and 5 foxes km -1 was calculated for the transects. Overall, 15/30 (50%) of all recovered foxes had not been detected by scat genotyping, 23/53 (49%) of KTBA genotypes were detected only once and 5/54 (9.5%) of foxes were found to have moved between two transects.Conclusions. At transects where population manipulation occurred, surviving individuals contributed significantly more scats than at the control transects and some individuals were detected at bait stations at a much greater frequency. This strongly suggested that they had contributed disproportionately to some IMI density estimates that were probably influenced by a change in the activity of some individuals rather than changes in population density alone. At one transect, eight foxes were confirmed to be present by spotlight surveys and were detected by scat and KTBA genotypes, yet were undetected by scat, bait station and sand plot indices.Implications. Scat and other DNA-based survey techniques provide a great deal of information about the identification and movement of individuals and if DNA sampling methods can be made more efficient they have the potential to provide accurate abundance estimates that are independent of the control technique.

Carnivorous predators are difficult to detect using conventional survey methods, especially at low levels of abundance. The introduced red fox (Vulpes vulpes) in Australia is monitored to determine the effectiveness of control programs, but assessing population parameters such as abundance and recruitment is difficult. We carried out a feasibility study to determine the effectiveness of using faecal DNA analysis methods to identify individual foxes and to assess abundance before and after lethal control. Fox faeces were collected in two sampling periods over four separate transects, and genotyped at five microsatellite loci. Two transects were subject to lethal control between collection periods. DNA was extracted from 170 fox faeces and, in total, 54 unique genotypes were identified. Fifteen biopsy genotypes from 30 foxes killed during lethal control were detected among the faecal genotypes. Overall, a similar number of genotypes were detected in both sampling periods. The number of individuals sampled in both periods was low (n = 6) and new individuals (n = 24) were detected in the second collection period. We were also able to detect animals that avoided lethal control, and movement of individuals between transects. The ability to identify individual foxes using these DNA techniques highlighted the shortcomings of the sample design, in particular the spatial scale and distances between transects. This study shows that noninvasive DNA sampling can provide valuable insight into pre and post fox abundance in relation to lethal control, individual behaviour and movement, as well as sample design. The information gained from this study will contribute to the design of future studies and, ultimately, control strategies.

Fluoroacetic acid (1080) is frequently used to poison the introduced red fox (Vulpes vulpes) in Australia. The symptoms of 1080 poisoning in dogs appear extremely distressing to observers as manic running, yelping and convulsing are readily interpreted as being indicative of pain and distress. Assessment of pain perceived by animals poisoned by 1080 is difficult, as severe CNS disruptions alter behaviour and EEG patterns that may otherwise be useful in such assessments. This study compared three drug agents combined with 1080 to address the possibility of pain and distress that may be experienced by foxes during 1080 toxicosis. A mixed-sex group of 15 foxes was used in the trial of each of the three drugs: an analgesic (carprofen; 10 mg kg–1), an anxiolytic/sedative (diazepam; 10 mg kg–1) or an analgesic/sedative (clonidine; 0.75 mg kg–1). Each group of 15 foxes was randomly allocated between three treatments of either 0.5 mg kg–1 of 1080, 0.5 mg kg–1 of 1080 and a dose of the trial drug, and a dose of the trial drug alone. A telemetry collar was used on each fox to monitor the duration and intensity of activity, and behaviour from dosage to death was recorded using daylight/infra-red video cameras. In foxes dosed with 1080 alone, a mean of 4.05 (0.86, P < 0.05) h was observed from dosage to symptoms and 1.57 (0.46, P < 0.05) h from symptoms to death. Diazepam reduced the overall intensity of activity from dosage to death (P < 0.002) and from first symptoms to death (P < 0.05). Diazepam also extended the time until death (P < 0.01) and time taken from dosage to first symptoms (P < 0.01). In the dosages used, clonidine was not as effective as diazepam in reducing activity associated with symptoms of poisoning, although it prolonged the time taken from dosing to first symptoms (P < 0.05). Carprofen did not significantly alter the progression of 1080 toxicosis or the intensity of activity of foxes compared with the group that received 1080 alone. The initial symptoms of 1080 toxicosis include retching and manic running and foxes may be more capable of suffering at this stage than after collapse, where behaviour is likely to be associated with CNS disruption. Diazepam may be effective in minimising any anxiety experienced by foxes especially during the first symptoms of 1080 toxicosis.

Predation by feral cats (Felis catus) and red foxes ( Vulpes vulpes) are key threatening processes for many endangered wildlife species. Toxin delivery through compulsive oral grooming is a potential mechanism to supplement existing control techniques for feral cats and red foxes, particularly when high prey densities reduce the uptake of toxic food baits by cats. We investigated the efficacy of different grooming traps by applying a gel containing toxic paraaminopropiophenone (PAPP) to the fur of feral cats and red foxes in experimental pens. Grooming behaviour and signs of poisoning in these animals were recorded by video. More cats interacted with "walk past" grooming traps triggered by sensor beams than with trap models that required the cat to enter a pipe or baited cage. After triggering a trap that had applied PAPP gel to their fur, 14 of 16 feral cats showed symptoms of anoxia, and 8 of these cats were dead by the following morning without exhibiting signs of distress. Seven of 12 foxes were observed to groom fur to which toxic gel had been applied and 3 of these ingested a lethal quantity of PAPP as a result. Our successful proof-of-concept trials support further development of grooming trap sensors and toxin delivery mechanisms to provide humane and targeted feral cat control, although this technique is unlikely to be as successful for fox control, given that foxes appear to not groom as fastidiously as cats.

Fox carcasses are seldom recovered after a 1080-baiting program, making analysis of efficacy difficult. Sodium cyanide was selected as an alternative toxin due to its rapid mode of action. A number of bait techniques were trialled in order to develop an appropriate cyanide delivery system that could be used as a buried bait. Techniques investigated included treated wax and gelatine capsules, a wooden capsule holder and a modified M-44 cyanide ejector. Subsequent trials showed that the modified M-44 ejector had greater efficacy in recovering fox carcasses at bait stations when compared with the other techniques trialled. This paper describes the range of baits trialled and the modifications that allow the M-44 to be used as a buried bait. A protocol for deployment of the M-44 in the field, together with a brief assessment of efficacy for each technique is also provided.