The feeding test results of cholecalciferol against wood rats, R. tiomanicus, showed that cholecalciferol is efficacious to control a major rat species in oil palm plantations compared to other ARs, particularly FGARs. Cholecalciferol is known to be effective against Norway rats (Rattus norvegicus) and house mice (Mus musculus). Marshall (1984) reported that all tested Norway rats (n = 145) and house mice (n = 100) died after consuming cholecalciferol baits with a 750 ppm (0.075%) concentration, with an average 3.9 to 6.1 days (mice) and 3.3 to 4.7 days (Norway rats) to result in death, though the number of days of the feeding trial was not mentioned in the report. Similarly, Eason et al. (2010) reported that cholecalciferol baits at a higher concentration at 0.8% (originally used to control possums, Trichosurus vulpecula) was effective against Norway rats and house rats (Rattus rattus). The authors ran a feeding trial of choice and no-choice feeding for two days on 35 rats consisting of ship rats (n = 20) and Norway rats (n = 15). Mortality was recorded in 34 of the 35 rats (97%) in an average of four days.
In our study, cholecalciferol produced similar results as chlorophacinone against wood rats, i.e., mortality rate and days-to-death, despite cholecalciferol having a lower daily dosage consumption than chlorophacinone. Based on the oral LD50 of both compounds on Norway rats, LD50 of cholecalciferol (43.6 mg/kg) was higher than chlorophacinone (20.5 mg/kg) (Marshall, 1984; PMEP, 2001; Eason et al., 2010). This shows that the rats are more susceptible to chlorophacinone than cholecalciferol due to the lower LD50 value. Hence, the high concentration of cholecalciferol in formulated baits compared to other anticoagulants was necessary since only a high concentration of the compound can make the formulated rodenticide effective against target pests (Kaukeinen et al. 2000). Based on a publication by Lund (1971), one day feeding of chlorophacinone with concentration of 0.025% was able to result in 60% mortality against mice (n = 20) and the mortality achieved was more than 70% after two days feeding of the bait. 95% mortality was achieved after 21 days of multiple feeding against the mice. According to publication results by Marshall (1984) mentioned earlier, all tested mice (n = 100) died after fed baits containing 0.075% cholecalciferol. However, a contradiction to the results reported by Marshall (1984) was reported by Hix (2009a) where 0.075% of cholecalciferol was only sufficient to kill rats but not mice. Hix (2009b) stated that based on feeding trials, both rat and mice can be controlled with higher dosage at 0.4% cholecalciferol.
The poor performance by warfarin compared to cholecalciferol was expected as reflected in the mortality rate, as the mortality of rats in the warfarin treatment did not even reach half of the total sample of rats tested. Even though the amount of bait consumed per day by rats in the feeding trial was the highest among all treatments, it only resulted in 44% of mortality of the rat samples, though warfarin took a shorter time than cholecalciferol to result in mortality of rats. The poor performance of warfarin has been reported in many publications due to resistance which was recorded since the 1980s against wood rats in Malaysia (Wood & Liau, 1977; Wood & Chung, 1990; 2003). However, there have been no proper laboratory trials conducted against wood rats except by Lee and Mustafa (1983). The researchers reported that warfarin compound was effective against 80–95% of wood rat samples (n = 20) but it took 6–8 days feeding of the baits. This showed that warfarin was still effective against wood rats at that time, as the initial reports of wood rat resistance against warfarin was only documented in 1983 and began to be considered a serious problem after 1985 when three different localities in Malaysia had the same resistance problem (Wood & Chung, 1990). The results of Lee and Mustafa (1983) are in contrast with our current study where the rats consumed 31.10 ± 1.96 mg/kg a day for 6 days feeding but were unable to achieve 50% mortality; indicating that the species can tolerate the toxin compared to the situation of 36 years ago. There is no species-specific LD50 of warfarin against wood rats, but the information is available for house mice and Norway rats. A higher acute oral LD50 of warfarin is reported for house mice at 374 mg/kg, while for Norway rats it is between 58–323 mg/kg, depending on the strain (Hagan & Radomski, 1953; Erickson & Urban, 2004). In comparison, cholecalciferol has a lower LD50 recorded against mice and Norway rats at 42.5 and 43.6 mg/kg respectively, reflecting that rodents are more susceptible to cholecalciferol compared to warfarin.
In Malaysia, secondary poisoning is one of the main issues with regards to AR usage since in oil palm plantations (one of the main agriculture sectors in Malaysia) barn owls are utilized as a biological control agent in order to reduce dependency on chemical practices such as rodenticides (Wood & Chung, 2003). Uncontrolled application of ARs can result in deleterious effects on barn owls as the diet of the owls are highly dependent on rats which make up about 99% of prey (Lenton, 1984; Hafidzi et al., 1999; Salim et al., 2014). There are various reports on secondary poisoning risks of ARs against non-target animals, including barn owls (Tyto alba). However, there is a lack of information on the effects of cholecalciferol on barn owls, though the compound is considered less toxic to bird species which appear to be less sensitive than mammals from the species tested in the literature (Eason et al., 2000; Erickson & Urban, 2004; Eason & Ogilvie, 2009). Marshall (1984) stated that oral LD50 of mallard duck was as high as 2000 mg/kg while Eason et al. (2000) proved that even at 2000 mg/kg dosing, no deaths were recorded out of the 6 ducks tested. On the other hand, Eason et al. (2000) stated that 3 out of 4 and 1 out of 4 of total domestic chickens and canaries tested died when given the same dose. Meanwhile in the same study, 10 of 16 weka (Gallirallus australis) which voluntarily ingested 0.1% cholecalciferol bait exhibited no signs of toxicity.
In addition to lower primary toxicity of cholecalciferol as reported in the aforementioned acute toxicity bird studies, there are published secondary toxicity studies of cholecalciferol towards a range of animals. Eason et al. (1996, 2000) reported that feral cats showed no signs of poisoning after being fed with poisoned-possum carcasses for five to six days. However, dogs are quite susceptible to cholecalciferol regardless of primary or secondary exposure. Gunther et al. (1988) stated that all four tested dogs died after two dogs consumed a high dose of 540g bait (20 mg ai/kg) or equal to one fourth of LD50 while the other two were given half of that amount, 10 mg ai/kg or equal to one eighth of the LD50. The author also reported that all four dogs developed signs of poisoning before death such as lethargy, weakness and anorexia within 48 hours and recorded death after 65 to 77 hours of treatment. Furthermore, the paper also mentioned that further investigation reported that all four dogs developed hypercalcemia and hyperphosphatemia which was concluded through detection of gastrointestinal hemorrhage, myocardial necrosis and mineralization of vascular walls.
Secondary exposure toxicity to dogs was reported by Eason et al. (2000) where the signs of toxicity detected on tested dogs fed on five poisoned-possum carcasses from acute group (the possum were dosed prior to the feeding period with LD95 dose of 0.8% cholecalciferol-treated cereal bait before humanely euthanized after 48 hours of dosing) were observed having partial anorexia and lethargy after 4 to 14 days of feeding. Despite being affected by the treatment, the author also reported that all affected dogs recovered gradually after about 14 days of exposure with the appetite and body weight returning to pre-treatment conditions without any veterinary intervention. According to similar study but against lower concentration of cholecalciferol (commonly used as rodenticide) conducted by Marshall (1984) there were no sign of toxicosis observed on six beagle-type dogs after being fed with poisoned-carcasses of Norway rats, which died after fed on 0.075% cholecalciferol bait prior to the feeding trial. In our present study, we let the rats voluntarily consume the 0.075% cholecalciferol bait for two days and recorded an average 0.08 mg/g a.i per body weight. This is far lower than the dose given by Eason et al. (2000) to the possums before feeding them to cat and dogs but similar to concentration used by Marshall (1984) because in this study we used cholecalciferol bait for rodents (0.075% a.i.) where the a.i concentration was ten times lower than possum baits (0.8% a.i.). The barn owls which consumed the poisoned rats did not display any signs of toxicity from secondarily consuming cholecalciferol such as typical behaviour aberration. For instance, the behaviours that are commonly observed in poisoned barn owls are less flying activity, passive manner and spending more time on the ground rather than on perching point, along with reduction of weight as stated by Hasber et al(2014) as symptoms of barn owls affected due to secondary poisoning of AR rodenticides.
As mentioned above, there have been assessments of secondary poisoning risks of several ARs against barn owls. Mendenhall and Pank (1980) stated that consumption of rats poisoned with SGAR compounds such as bromadiolone, brodifacoum and difenacoum by barn owls caused hemorrhaging effect after a one-week feeding trial. Gray et al. (1994) recorded that at least one out of four barn owls did not survive the feeding test of mice poisoned with brodifacoum and difenacoum while two of four owls did not survive after consuming flocoumafen-poisoned mice. In Malaysia, Lee (1994) stated that not only SGARs such as bromadiolone, brodifacoum and flocoumafen, but FGAR compounds such as warfarin also caused high degree of toxic effects on barn owls. The author fed four barn owls with poison rats treated with SGAR bait which resulting in death of three of four owl samples after two weeks of the exposure, while in the same study against another group of barn owls (n = 4) fed with poison rats treated with FGAR compound, warfarin have caused death on two of four tested barn owls after three weeks of exposure to the poisoned rats. A study conducted by Salim et al. (2014) against two group of four barn owls where each group of barn owls were fed with chlorophacinone and bromadiolone treated rats. The result showed that one of four tested owl samples from each group were observed with following sign of poisoning, coarse breathing, reduce of weight and flying activity, hemorrhage at the beak and hematoma (bromadiolone) after consumed three poisoned rats in a week of feeding period.
The secondary toxicity effect of ARs is not only reported in barn owls, but also in other non-target raptors. Lutz (1986) recorded an increase in blood coagulation after Eurasian buzzards (Buteo buteo) were fed with bromadiolone-poisoned mice for 10 days. Grolleau et al (1989) observed that 27 Eurasian buzzards (Buteo buteo) exhibited bleeding after feeding on bromadiolone-poisoned voles for three days. In the same study with mammals, 10 tested ermines (Mustela erminea) were observed to be bleeding after being fed with bromadiolone-poisoned vole for three to five days. Twenty American kestrels exhibited external bleeding after feeding on chlorophacinone-poisoned voles for one to three days (Radvanyi et al., 1988) while an increase in blood coagulation time was reported in Eurasion buzzards after fed poisoned mice (Riedel et al., 1991). Several predators such as black kites, red kites, short toed snake-eagles, and golden eagles showed flocoumafen contamination, as reported by Sanchez-Barbudo et al. (2012) in an opportunistic study on carcasses. Warfarin is generally considered less hazardous to non-target animals. Minks, least weasels and dogs have recorded deaths and survivors exhibited external bleeding signs after eating prey poisoned with warfarin (Prier & Derse, 1962; Evan & Ward, 1967; Townsend et al., 1984).
One of the reasons our secondary poisoning assessment on barn owls did not include other ARs used in the feeding trial was due to the fact that data on the poisoning of barn owls are already established from past publications. Moreover, barn owls are a protected species in Malaysia under Act 716, Wildlife Conservation Act 2010. Thus, only a limited number of samples were permitted for this study and there was no necessity to run a higher number of samples simply to confirm known poisoning effects from other ARs. Past publications such as those mentioned above, have reported the toxic effects of ARs, both FGARs and SGARs, on non-target barn owls via secondary poisoning.