- Email: [email protected]
Rodenticides in British Barn Owls
Environmental Pollution 68 (1990) 101-117
Rodenticides in British Barn O w l s
I. Newton, I. Wyllie & P. Freestone Institute of Terrestrial Ecology, Monks Wood Experimental Station,
Abbots Ripton, Huntingdon, Cambs PE17 2LS, UK (Received 14 December 1989; revised version received 4 July 1990; accepted 6 July 1990)
ABSTRACT Out of 145 Barn Owls found dead through accidents (66%), starvation (32%), shooting ( 2 % ) and poisoning ( < 1% ), 10% contained residues of rodenticides, difenacoum or brodifacoum, in their livers. Difenacoum was in the range 0"005q9"106 I~gg- l fresh weight, and brodifacoum was in the range 0"019-0"515 I~gg - ~. Minimum levels of detection were about 0"005 #g g - 1.for both chemicals. Mice fed for 1 day on food containing difenacoum and brodifacoum died after 2-11 days. Within these mice residues were present at greater concentration in the liver than in the rest of the carcass. The mean mass of residue in a whole 35g mouse was estimated at lO'17t~g (range 4"7320"65 l~g) for difenacoum and 15"36 pg (range 8.07-26.55)for brodifacoum. Such poisoned mice were fed to Barn Owls for successive periods of 1, 3 and 6 days. All six owls fed on difenacoum-dosed mice survived all three treatments, in which up to an estimated lO1.7#g of difenacoum was consumed, and the coagulation times of their blood returned to near normal in less than 5 23 days. Four of the six owls fed on brodifacoum-dosed mice died 6-17 days after the 1-day treatment, but the survivors also survived the 3-day and 6-day treatments. Those that died had each eaten 3 mice, with a combined weight of about 105 g and a total brodifacoum content of about 46.07#g, which was equivalent to a dose of O'15Oqg.182 mg kg-1 of owl body weight. After death these owls had 0"63-1"25 i~gg-x of brodifacoum in their livers. Blood from the survivors would not coagulate at 9 days post-treatment, but did so at 16 days in one bird and between 38 and 78 days in the other. It is concluded that: (1) Barn Owls in Britain are now widely exposed to second-generation rodenticides; (2) not all owls exposed to these chemicals 101 Environ. Pollut. 0269-7491/90/$03.50 © 1990 Elsevier Science Publishers Ltd, England. Printed in Great Britain
102
I. Newton, L Wyllie, P. Freestone are likely to receive a lethal dose; (3) brodifacoum is more toxic to owls than difenacoum; and (4) while there is yet no evidence that rodenticides have had an), appreciable effect on Barn Owl populations in Britain,further monitoring of residue levels and population trends is desirable.
INTRODUCTION This paper has two aims. The first is to report the incidence of two rodenticides, difenacoum and brodifacoum, in the bodies of Barn Owls Tyto alba found dead in various parts of Britain. The second is to report the effects of feeding captive Barn Owls on mice killed by these two rodenticides. The chemicals concerned are two of the second-generation rodenticides, developed as replacements to warfarin and other first generation rodenticides, to which rodents in some regions have become resistant. Like warfarin, both chemicals are coumarin derivatives (based on the 4-hydroxy coumarin moeity) and they act as anti-coagulants. They are more toxic and persistent than warfarin, however, and have been reported to cause secondary poisoning in various rodent-predators, including owls (Mendenhall & Pank, 1980, Hegdal & Colvin, 1988). The work described here was intended to help in assessing the impact (if any) of these new rodenticides on Barn Owls in Britain, which are thought to have declined in numbers in recent decades (Bunn et al., 1982; Shawyer, 1987). The chemical name of difenacoum is given as 3-(3-biphenyl-4-yl1,2,3,4-tetrahydro-l-naphthyl)-4-hydroxycoumarin, and of brodifacoum as 3- [3-(4'-bromobiphenyl-4-yl)- 1,2,3,4-tetrahydro- 1-naphthyl]-4-hydroxycoumarin (Worthing & Walker, 1987). Difenacoum has been available in Britain since 1975, and has been used both indoors and outdoors. Brodifacoum has been available since 1982, but apart from some experimental outdoor trials, its use has been restricted to within buildings by professional operators. Both chemicals have been commonly used around farmsteads and other sites, where rodents are a problem. As farm buildings provide some of the most frequently used roosts and nest-sites for Barn Owls in Britain, these owls would be expected to be at risk, as they eat affected rats and mice. Brown et al. (1988) listed five Barn Owls among various rodenticide victims examined during 1973-86. Likewise, Shawyer (1987) listed four Barn Owls found dead or dying, associated with local difenacoum use, and four others following brodifacoum use; all showed symptoms of haemorrhaging. He questioned whether these chemicals may have contributed to a decline in Barn Owl numbers in parts of Britain. In a more extreme situation, Duckett (1984) attributed the collapse of a Barn Owl population in a Malayan oil palm plantation to use of
Rodenticides in British barn owls
103
brodifacoum and coumachlor against rats which formed the main prey. Meanwhile, the high toxicity of brodifacoum to wild or captive owls of various species has been confirmed by Mendenhall & Pank (1980), Merson et al. (1984) and Hegdal and Colvin (1988). To our knowledge, however, there has been no previous survey of the incidence of rodenticides in wild Barn Owls and no measurement of tissue levels in Barn Owls known to have died of rodenticide poisoning. METHODS
Survey Advertisements were placed in various bird journals asking for the bodies of Barn Owls found dead. All carcasses were requested, irrespective of mode of death. On receipt, the bodies were stored deep frozen until they could be examined, usually some months later. After thawing, a post-mortem was conducted by visual inspection only. Where possible, the findings were used, along with information from the sender, to determine the cause of death. Careful search was made for signs of haemorrhaging, which often follow an exposure to anticoagulants. Typical sites for rodenticide-induced haemorrhaging include the muscle on both sides of the breast bone, the leg joints, the m o u t h and the nares, as well as various internal organs. After inspection, part of the liver was removed from each bird and analysed for residues of difenacoum and brodifacoum. The method of analysis was that of Hunter (1985) modified in minor respects. Liver samples were extracted with chloroform-acetone, and the extracts were cleared of fat using Bond-Elut columns. The concentrated samples from the columns were then analysed by Varian High Pressure Liquid Chromatography (HPLC), using a M C H - 5 N micropack column and spectrofluorometer, against a standard for each c o m p o u n d . When difenacoum or brodifacoum was detected, a recovery test was done from a spiked sample of solvent to validate the identification and to correct the estimate of mass present. Recoveries from most batches were in the range 75-95%. Specimens in which no residues were detected gave blank chromatograph traces, showing that no other material with similar properties was present. In addition, however, confirmation of the chemical identity of residues in all birds in which residues were detected was obtained using an alternative H P L C column and solvents. The lower limit of detection for both compounds was around 0.01 pg, which was equivalent to 0.005-0.01 pg g - 1, depending on the weight of the sample. For calculation of average concentrations in known contaminated
104
I. Newton, L Willie, P. Freestone
mice, nil detected values were taken as 0.0075/zg g - 1. Both isomers of each c o m p o u n d are included together in the figures reported here. Concentrations are given as wet weight values, but can be converted to dry weight values by multiplying by 3.64 for liver of owls, by 4.30 for liver of poisoned mice and by 3"29 for whole carcasses of poisoned mice (see later). The values from wild owls were corrected to allow for water loss, assuming a wet to dry conversion factor in fresh liver of 3.64, the value found for captive owls.
Toxicity trials Owls were dosed with rodenticide-poisoned mice. The owls were kept individually in cages measuring 60 x 50 × 40 cm high, and were thus much less active than wild owls, and were fed on white laboratory mice, rather than on wild rodents. As owls which are fed on fresh carcasses do not need to drink, no water was provided. Some days before a feeding trial, mice were fed for 1 day (no choice) on a proprietary dosed-food mixture containing either 0.005% difenacoum or 0.002% brodifacoum. After dying, some dats later, most mice were fed to owls. The mean weight ( ___SE) of the dead mice was 34.7 _ 1.1 g. Other dead mice, from the same batches, were analysed to determine the total content of rodenticide in their bodies. The liver was analysed separately from the rest of the carcass, which was first homogenised by crushing with liquid nitrogen. All the owls were captive-bred (second or third generation in captivity), very tame, and in their first year of life. They had not previously been exposed to rodenticides, and at the start of the trials in September they were in good condition, weighing 276-370 g. They were allocated at random into two groups of six, one fed on difenacoum-treated mice and the other on brodifacoum-treated mice. Owls in each group were fed initially for 1 day on dosed mice (3 per owl)~ Those which survived this treatment were later fed for three successive days on dosed mice, and any which survived this treatment were later fed for 6 days on dosed mice. The livers of owls which died were analysed for rodenticide residues in the same way as the livers from wild owls. Before the trials, and at various stages during it, small blood samples were taken from the brachial vein in some of the owls. In each case the coagulation time was measured, while tipping the blood back and forth in a 75 m m glass capillary tube of 1 m m internal diameter. The tube contained a 1 cm metal rod (a 'flea'), and coagulation time was taken when the rod would no longer slide around within the blood. All such sampling was done at the same time of day, at 1600 h. However, in view of the risks of haemorrhaging,
Rodent&ides in British barn owls
105
through blood-sampling birds treated with anticoagulant, such sampling was kept to a minimum. No problems were encountered, as the vein closed itself soon after the needle was withdrawn. With such a crude technique little significance could be attached to small variations in coagulation times; the main purpose of the test was to check whether the blood would coagulate or not. The findings were used mainly to indicate when recovery from exposure to the rodenticide occurred, and when another series of trials could be started. RESULTS
Survey In the period 1983-89, a total of 145 Barn Owls was received from various parts of Britain for analysis. On visual inspection of the opened carcass, together with information from the sender, most were diagnosed as victims of accidents or starvation (Table 1). In only one were the typical haemorrhage symptoms of a rodenticide victim noted. On later chemical anlysis, difenacoum was detected (and confirmed) in the livers of seven individuals, brodifacoum in four, and both chemicals together in another four, making a total of 15 (10%) contaminated birds (Table 2). These birds came from various parts of Britain (Fig. 1), but about half seemed to be from warfarin-resistance areas, as depicted by Shawyer (1987). No similar chemicals, such as bromadiolone and flocoumafen, were detected. There was no obvious seasonal variation in the percentage of received carcasses that contained rodenticide and no obvious sex bias, as TABLE 1 Causes of Death Diagnosed on Post-mortem Examination of 145 Wild Barn Owls. 1983-89
Trauma (road accident) Trauma (other accident)a Starvation Shot Poisoned
Number
Number with rodenticide residues
81 14 46 3 1
6 4 4 0 1
a Includes mainly collision victims, two drowned and one electrocuted.
TABLE 2 Numbers of Wild Barn Owls Tested for Rodenticides each Year from 1983 Showing Numbers Containing Brodifacoum and Difenacoum
Year
Number tested
Difenacoum only
Brodifacoum only
Both rodenticides
% with detectable residues
1983 1984 1985 1986 1987 1988 1989= Overall
3 15 31 36 22 23 15 145
0 0 1 2 2 1 1 7
0 1 1 0 0 1 1 4
0 0 1 2 0 1 0 4
0 7 10 11 9 13 13 10
" To 31 March only.
• Brodifacoum • Difenacoum =,,Both c h e m i c a l s o None detected
¢
F
0 ~
0-~
o
~
og "o
~o
7
%
o
oo
0
0 0 O0
'o o
o
o
g~ ° o o
%L°° °. o
Fig. 1.
o o
Map showing locations of Barn Owls found dead and received for analysis.
Rodenticides in British barn owls
107
1 0 % o f m a l e s h a d residues a n d 1 1 % o f females. H o w e v e r , o n l y 7 % o f firsty e a r b i r d s h a d d e t e c t a b l e residues, c o m p a r e d w i t h 15 % o f o l d e r birds. Levels o f d i f e n a c o u m w e r e in the r a n g e 0 . 0 0 5 - 0 - 1 0 6 p g g -1 ( m e a n 0.036 # g g - 1 ) , whilst levels o f b r o d i f a c o u m w e r e in the r a n g e 0"019-0"515/~gg -~ ( m e a n 0 - 1 8 0 # g g - x ) ( T a b l e 3). T h e single b i r d d i a g n o s e d o n p o s t - m o r t e m as a r o d e n t i c i d e victim h a d 0.43/~g g - a b r o d i f a c o u m in its liver.
Toxicity trials: Mice All d o s e d m i c e died 2-11 ( m o s t l y 3-8) d a y s a f t e r dosing, w i t h n o significant difference b e t w e e n d i f e n a c o u m a n d b r o d i f a c o u m - t r e a t e d a n i m a l s (Fig. 2). C h e m i c a l a n a l y s i s o f a s u b s a m p l e s h o w e d that, f o r b o t h chemicals, the residue w a s m u c h m o r e c o n c e n t r a t e d in the liver t h a n in the rest o f the
TABLE 3 Levels (#gg ~ in Corrected Wet Weight a) of Rodenticides found in the Livers of Contaminated Barn Owls Specimen number
Date of finding
Location
Age
Sex
Levels of rodenticide Difenacoum
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
16.11.84 12.2.85 22.5.85 31.5.85 16.1.86 9.8.86 4.9.86 20.11.86 12.5.87 13.11.87 12.1.88 8.2.88 2.12.88 --.2.89 15.3.89
Hants. Notts. Kent W. Sussex Powys Wilts. Humberside Isle of Man Beds. Salop Dyfed Suffolk Norfolk Wilts. Gwynedd
Y A A A Y A A Y A A Y A A Y A
F M F F F M F M M F M F F M M
ND ND 0"052 0-024 0.010 0.051 0"019 0-106 0"024 0-005b 0-061 ND 0-034 0.012 ND
Brodifacoum 0.515 0-300 ND 0.035 ND 0.035 ND 0"070 ND ND 0-040 0-423 ND ND 0-019
A = Adult. Y = First year bird. F = Female. M = Male. a Levels were obtained on a dry weight basis and converted to wet weight values by dividing them by 3.64, the mean wet to dry ratio in fresh owl liver. This procedure was necessary because some of the specimens had dehydrated to some extent. b Close to the limit of detection. ND, not detected.
L Newton, 1. Wyllie, P. Freestone
108
P e r c e n t a g e d e a t h s of d o s e d m i c e p e r day a f t e r dosing 50
40 :~
30
.~
20
I
Brodifacoum(N:41) D i f e n a c o u m (N=113)
I 10 °
clays a f t e r d o s a g e
Fig. 2.
Time to death in mice dosed with difenacoum and brodifacoum.
carcass (Table 4). However, because the liver forms only about one-fifteenth of the weight of a mouse, the total mass of rodenticide in the rest of the carcass was substantial, especially for brodifacoum (Table 4). On the basis of our results, a 35 g mouse on the day of death contained a total of 10.17 pg of difenacoum, on average, or a total of 15.36 pg of brodifacoum.
Toxicity trials: Owls All of the owls fed on difenacoum-dosed mice survived the 1-day, 3-day and 6-day treatments (Table 5), and none showed external bleeding. After the 1day treatment, all six owls were blood-sampled 5-9 days later, and coagulation times were 'normal' (Table 6). After the 3-day treatment, the blood of one bird sampled 3 days later would not coagulate, even after 24 h, but all birds seemed to have become normal in this respect 9-23 days after treatment. With the method used, no significance could be given to the fact that, in four of the six birds, coagulation times were somewhat shorter after recovery from difenacoum than before exposure to it (Table 6). Of the six owls fed on brodifacoum-dosed mice, four (one male and three females) died 6, 10, 11 and 17 days after the 1-day treatment, with 0"631.25 pg g - 1 brodifacoum in their livers (Table 5). The weight changes in three of these birds between treatment and death were small ( + 1%, 0%, - 6%), but the fourth bird (which died 10 days post-treatment) lost 23%, declining from 332 g to 257 g. On post-mortem, all showed internal bleeding: down
109
Rodenticides in British barn owls
+l "4D
.=_
¥, oo
¢'4
+L ()
6~
<:
+l
0
0
.<
6 ~ 6 6 6 ~ 6 ~ 6 ~ .=_ =_
÷~ r'0
C)
.>_
4r
_
.=.
~r
.o E
m
()
0
()
÷l
["
z
110
I. Newton, L Wyllie, P. Freestone
TABLE 5 Results of Feeding Rodenticide-poisoned Mice to Barn Owls
Difenaeoum
Brodifaeoum
Period o f dosing (days)
Mice eaten
Number o f owls dosed
Number surviving
1 3 6 1 3 6
3 6 5-10 3 6 12
6 6 6 6 2 2
6 6 6 2 2 2
Notes: The brodifacoum levels in the livers of the 4 owls which died were calculated at 0-63, 0-77, 1.02 and 1'25/~g g-~. The gaps between the start of successive dosing periods were 11-14 and 33-35 days for difenacoum, and 77-79 and 75 days for brodifacoum.
both sides of the keel bone, and around the brain, heart, lungs or gut, so that in two birds blood filled the body cavity. The two surviving owls also survived the later 3-day and 6-day treatments with brodifacoum. Again, however, they were affected, as they showed prolonged bleeding from the mouth or feet for up to 30 days post-treatment. Blood taken from two birds 9 days after treatment would not coagulate, nor would blood taken from one of these birds (a survivor) after 38 days. However, blood from the other survivor seemed normal 16 days after treatment (Table 6). It seems, therefore, that brodifacoum is more toxic than difenacoum to Barn Owls, and in survivors, has much more prolonged effects. From knowledge of the number of mice eaten by the owls, and the average weight of rodenticide in the mouse bodies at the time, the total mass of rodenticide consumed by the owls in the 1-, 3- and 6-day treatments could be calculated (Table 7). Individuals consumed an estimated 30.5 #g, 61.0 #g and 50-9-101"7 #g difenacoum in 1, 3 and 6 days, respectively, and an estimated 46.1 #g, 92.2 #g and 184.3 #g brodifacoum. The 46.1 #g 1-day dose of brodifacoum was enough to kill four of the six owls, and was equivalent to a dose of 0"150-0"182mgkg -1 body weight for the birds concerned. However, considering the variability in residue levels in individual mice (Table 4), these figures should be regarded as no more than rough approximations. The analysis of regurgitated pellets (of fur and other undigested parts of food items) from each group of owls revealed that some of the rodenticide that was consumed was excreted unchanged in pellets. In other words, it had not passed through the gut. Five pellets collected from one group of owls
Rodenticides in British barn owls
111
TABLE 6 Coagulation Times (Minutes, Seconds) for Barn Owl Blood Samples taken on Different Dates after End of Dosing
Dijenacoum
Owl no. 1
2
3
4
5
6
Pre-treatment
2, 00
2, 30
4, 45
4, 15
2, 45
1, 45
After 5 7 8 9
1-day dose days days days days
. --2, 30
-3, 45 --
1, 30 ---
After 3 9 10 21
3-day dose days days days days
22 d a y s 23 d a y s
NC . --
.
.
.
-1, 30 --
1, 45
-.
-1, 30
. 6, 15 --
-. . --
1, 30 --
1, 45 --
Brodifacoum 7
8
1, 15
1, 45
A f t e r 1-day d o s e 9 days 16 d a y s
NC --
NC" --
38 d a y s 78 d a y s After 3-day dose 60 d a y s
-. .
5, 00 . 2, 30
. 6, 00
2, 00
--
Owl no.
Pre-treatment
24 d a y s 36 d a y s
1, 00 ----
. .
. .
9 1, 15 a ---
. .
NC 4, 30
. .
5, 30
--
10
11
12
3, 30"
4, 00
1, 15"
--
-6, O0
. . . .
--
8, 30 5, 30 . .
. . --
. . 0, 15
--
a Died before further blood samples could be taken. N C = N o c o a g u l a t i o n w i t h i n 24 h.
during the feeding trial contained a mean of 0.60 pg difenacoum (range 0-06-1-19#g), whilst six pellets from the other group contained a mean of 0.97 pg brodifacoum (range 0"08-3.86 #g). Assuming a regurgitation rate of two pellets per day, this would have reduced the total rodenticide consumption to 29.3/~g, 57"4 pg and 43.7-94.5 pg in the 1-, 3- and 6-day difenacoum trial, and to 44.1 #g, 86.4 pg and 172.8 #g in the 1-, 3- and 6-day brodifacoum trial. Again these estimates can only be rough, because of the great variation in the rodenticide content of particular pellets.
112
L Newton, L Wyllie, P. Freestone
TABLE 7 Calculation of Mass of Difenacoum and Brodifacoum Consumed by Owls in the Feeding Trials Days
Difenacoum
Brodifacoum
of dosing
1 3 6
Number
Mean mass
Total mass
Number
Mean mass
of
of
of
of
of
of
mice eaten
difenacoum per mouse (lag)
difenacoum eaten per owl (lag)
mice eaten
brodifacoum per mouse (~g)
brodifacoum eaten per owl (lag)
30"51 61"02 50-85-101"71
3 6 12
15"36 15"36 15-36
46"08 92-16 184"32
3 6 5-10
10-17 10"17 10"17
Total mass
DISCUSSION
Survey About 10% of the 145 Barn Owls examined in 1983-89 had residues of difenacoum or brodifacoum in their bodies, but this sample may not reflect the true exposure of owls to these rodenticides in the areas concerned. Almost all the carcasses were found in the open, and were victims of road traffic and other accidents, or of apparent starvation. Death from rodenticide poisoning is delayed, and is preceded by lethargy, so poisoned owls might be most likely to die in their roosts, in roof-cavities or hollow trees, where they would be unlikely to be found by the casual observer. When they are found in such situations, moreover, they are seldom fresh enough for analysis. Any such bias in the sampling is entirely conjectural, however, and as yet there is no evidence one way or the other. Recent work on mammals and birds has shown the presence of binding sites for second-generation rodenticides in the liver, which enables residues to persist for up to several months after ingestion of a single oral dose (Parmar et al., 1987; Huckle et al., 1989a, b). In effect, this means that prey animals which took sub-lethal doses could remain as sources of contamination for owls for several months; and at the same time owls in which residues were detected could have been exposed up to several months previously or on more than one occasion. It is not certain, however, to what extent rodents acquire sublethal doses of these chemicals, as a single feed would normally be lethal. As they stand, our findings imply that the exposure of Barn Owls in Britain to difenacoum and brodifacoum is now frequent and widespread.
Rodenticides in British barn owls
113
Even if brodifacoum-use were restricted to the inside of buildings, as recommended, this would not wholly protect Barn Owls, which will sometimes hunt within buildings, and at the same time could take contaminated rodents which wander outside.
Toxicity trials The feeding trials on captive owls were conducted in order to help interpret the significance of residue levels found in wild owls. As our owls had been in captivity for only 1-2 generations, and were unrelated to one another, they could be considered as genetically representative of wild Barn Owls. Brodifacoum levels in the four captive birds that died were higher than those in the eight wild birds with brodifacoum. However, three of the wild owls had liver levels of brodifacoum approaching the lowest level in captive birds (0.30-0.52#gg-1 compared with 0.63pgg-l), while the five others had levels an order of magnitude lower (0"02-0.06 #g g- 1). There could have been at least three reasons for this difference. First, not all the wild owls may have died of rodenticide poisoning, but been exposed instead to sublethal levels. This explanation would be consistent with the low likelihood of poisoned owls being found (see above), and with the fact that, for most contaminated birds, other (non-rodenticide) causes of death were diagnosed on visual inspection. An alternative explanation would be that, owing to a behavioural difference, lower levels of rodenticides are needed to kill wild owls than captive ones. Our birds, kept in cages, remained still and inactive almost all the time, moving only when disturbed or to eat the mice provided. Yet those on brodifacoum still showed prolonged external bleeding. Wild birds, which must remain highly active, are presumably much more likely to suffer severe haemorrhaging from the exertion involved. Another likely limitation of our feeding trial is that it ran overwinter, when the owls were not moulting (apart from one which started during the 6-day brodifacoum trial). If the owls had been growing feathers throughout, the opportunities for haemorrhaging may have been much greater (though this did not occur in warfarin-fed moulting Tawny Owls Strix aluco, Townsend et al., 1981). A third explanation for the lower levels in wild than captive owls centres on the possibility of breakdown of residues after death. Although the highest residue was found in one of the longest-stored birds (4 years), the possibility of long-term post-mortem breakdown needs checking. Finally, there remains the possibility that sub-lethal levels of rodenticide may predispose death from other causes, or reduce the chance of recovery from accidents. On balance, however, we feel that the first explanation, that some wild owls received only sub-lethal levels of rodenticide and died from other causes, is the most likely reason why their levels were lower than those in captive owls.
114
1. Newton, L Wyllie, P. Freestone
On this basis, no more than 2% of the owls examined would have died of rodenticide poisoning. Despite these limitations, our results clearly imply that difenacoum is less toxic to Barn Owls than brodifacoum. None of the six birds fed difenacoum died, and blood coagulation was back to 'normal' within 5-9 days. In contrast, four of the six birds fed brodifacoum died within 6-17 days after a 1-day dose, and blood from survivors would still not coagulate at 9 days post-treatment. Birds that died ate three mice, with a combined weight of about 105 g and an estimated total brodifacoum content of 46.1 pg. The fact that two other birds of similar weight survived on the same 1-day intake, together with the later 3-day and 6-day treatments, could have been due to individual differences in sensitivity or behaviour (it was not related to sex). It was also possible that, owing to limited carrying capacity of the liver (Huckle et al., 1989b) or to continuing metabolism or excretion of residues, birds on 3- and 6-day treatments did not accumulate correspondingly greater concentrations in their tissues than did birds on the 1-day treatment. A third possibility is that birds which survived the 1-day treatment may have adjusted in some way, making them better able to withstand the 3- and 6-day treatments. These findings on the differential toxicity of the two compounds are consistent with those of Mendenhall and Pank (1980), who found that five out of six Barn Owls fed brodifacoum died, compared with none out of six fed difenacoum (but three bled externally). Nonetheless, deaths from difenacoum occurred in an extreme experiment with captive Tawny Owls Strix aluco, which were fed entirely on contaminated mice until they died after 8-41 days (Anon., 1982). The high toxicity of brodifacoum was confirmed in a field trial with voles and Screech Owls Otus asio in Virginia (Hegdal & Colvin, 1988). Five owls found dead 5-37 days post-treatment contained 0.4-0.8 pg g - 1 brodifacoum in their livers, whilst four caught and killed 34-43 days later contained 0.3-0.6 pg g-1. Two others in each group had no detectable residues. The levels in the dead Screech Owls were comparable to those in our Barn Owls of 0.63-1.25 pg g-1 suggesting a similar pre-death residue accumulation in the two species. Published values for the acute LDso for brodifacoum in different domesticated birds and mammals span two orders of magnitude: from 0.27 mg kg-1 body weight in the rat to about 25 mg kg-1 in the cat (Worthing & Walker, 1987). These figures imply a huge species variation in sensitivity. The estimated lethal dose of 0.15-0-18mgkg -~ brodifacoum which killed our owls was of similar order to the lowest published figures, making the Barn Owl one of the most sensitive of the species yet investigated. Our owls obtained this dose from three mice (total biomass 105 g), which falls within the range of 1-5 'rodents' estimated by Shawyer (1987) to be necessary to kill an owl.
Rodenticides in British barn owls
115
Normally, rodenticide residues in the rodent body reach a peak soon after ingestion, and then decline progressively until the animal dies some days later. Thus, if we had fed fresh mice to our owls, soon after bait ingestion, even fewer may have been needed to provide a lethal dose. Published figures for the acute LD 5o of difenacoum range between 0.8 and 100 mg kg-1 in different bird and mammal species, but for any one species the dose was higher than for brodifacoum (Worthing & Walker, 1987). Hence, the greater toxicity of brodifacoum over difenacoum, suggested by our results, was consistent with previous laboratory findings on other species. None of our owls would have received a dose of difenacoum as large as the smallest recorded LDso for this chemical (0.17mgkg -1 in owls compared with 0-8 mg kg- ~ in mice). In conclusion, second generation rodenticides (notably brodifacoum) could present a threat to Barn Owls and other predators of rodents. The several-day period between dosing and death in rodents, their lethargic behaviour for some hours before death, and the several-month persistence of sublethal residues within the mammalian body (Rammell et al., 1984), all facilitate the contamination of owls. Moreover, rodents that are commensal with man are widely eaten by Barn Owls, albeit in small amounts. House Mouse M u s musculus remains were present in 47% of 188 pellet samples examined by Glue (1974) from various parts of Britain, while Brown Rat Rattus norvegicus remains were present in 48% of samples. In a later Irish study, House Mouse remains were found in 10 out of 15 pellet samples, while Brown Rat remains appeared in 12 (Smal, 1987). The incidence of both species in the diet is generally small, but may well increase during hard weather or when voles are scarce. Field trials have given mixed results, but, where no owl mortality was recorded, this could be attributed to the radiomarked owls not hunting near baiting sites or taking mainly non-target prey (Kaukeinen, 1982; Hegdal & Blaskiewicz, 1984). Other field studies, in which the target rodent was the main prey, have shown mortality in owls (Merson et al., 1984, Hegdal & Colvin, 1988) and even substantial population decline (Duckett, 1984). As yet, however, despite widespread exposure and occasional mortality incidents, there is no firm evidence that modern rodenticides have had any appreciable influence on the population level of Barn Owls in Britain. However, more monitoring of residues and population trends is clearly desirable.
A C K N O W L E D G E M ENTS We are grateful to Dr P. Anderson, Dr S. Dobson and Mr C. Shawyer for useful advice at the outset, to Mr J. White for animal care, to Mrs A. Asher for help with data tabulation, and to A. P. Buckle, P. J. Edwards, S. Dobson,
116
L Newton, L Wyllie, P. Freestone
P. Greig-Smith, M. R. Hadler. To M. Roberts, C. Shawyer, I. R. Taylor and an a n o n y m o u s referee for helpful comments on the manuscript, The work was commissioned and funded by the Nature Conservancy Council, and the toxicity trials were undertaken under H o m e Office Licence. REFERENCES Anon. (1982). Secondary toxicity hazard to owls from difenacoum. In Pesticide Science 1981, Agricultural Science Service Research and Development Reports, MAFF, London, HMSO, pp. 36-7. Brown, R. A., Hardy, A. R., Greig-Smith, P. N. & Edwards, P. J. (1988). Assessing the impact of rodenticides on the environment. EPPO Bulletin, 18, 283-92. Bunn, D. S., Warburton, A. B. & Wilson, R. D. S. (1982). The Barn Owl. Calton, Poyser. Duckett, J. E. (1984). Barn Owls (Tyro alba) and the 'second generation' rat-baits utilised in oil palm plantations in Peninsular Malaysia. Planter, Kuala Lumpur, 60, 3-11. Glue, D. E. (1974). Food of the Barn Owl in Britain and Ireland. Bird Study 21, 200-10. Hegdal, P. L. & Blaskiewicz, R. W. (1984). Evaluation of the potential hazard to Barn Owls of talon (brodifacoum bait) used to control Rats and House Mice. Environ. Toxicol. & Chem., 3, 167-79. Hegdal, P. L. & Colvin, B. A. (1988). Potential hazard to Eastern Screech-owls and other raptors of brodifacoum bait used for vole control in orchards. Environ. Toxicol. & Chem., 7, 245-60. Huckle, K. R., Hutson, D. H., Logan, C. J., Morrison, B. J. & Warburton, P. A. (1989a). The fate of the rodenticide flocoumafen in the rat: Retention and elimination of a single oral dose. Pestic. Sci., 25, 297-312. Huckle, K. R., Warburton, P. A., Forbes, S. & Logan, C. J. (1989b). Studies on the fate of flocoumafen in the Japanese Quail (Coturnix coturnix japonica). Xenobiotica, 19, 51-62. Hunter, K. (1985). High-performance liquid chromatographic strategies for the determination and confirmation of anticoagulant rodenticide residues in animal tissues. J. Chromatography, 321, 255-72. Kaukeinen, D. (1982). A review of the secondary poisoning hazard potential to wildlife from the use of anticoagulent rodenticides. Proc. Vertebr. Pest. Conf., 10, 141-58. Mendenhall, V. M. & Pank, L. F. (1980). Secondary poisoning of owls by anticoagulant rodenticides. Wildl. Soc. Bull., 8, 311-15. Merson, M. H., Byers, R. E. & Kaukeinen, D. E. (1984). Residues of the rodenticide brodifacoum in voles and raptors after orchard treatment. J. Wildl. Manage., 48, 212 16. Parmar, G., Bratt, H., Moore, R. & Batten, P. L. (1987). Evidence for a common binding site in vivo for the retention of anticoagulants in rat liver. Human Toxicol., 6, 431. Rammell, C. G., Hoogenboom, J. J. L., Cotter, M., Williams, J. M. & Bell, J. (1984). Brodifacoum residues in target and non-target animals following rabbit poisoning trials. N.Z.J. Exp. Agric., 12, 107-11.
Rodenticides in British barn owls
117
Shawyer, C. R. (1987). The Barn Owl in the British Isles. Its past, present and future. London, The Hawk Trust. Smal, C. M. (1987). The diet of the Barn Owl Tyto alba in southern Ireland, with reference to a recently introduced prey species--the Bank Vole Clethrionomys glareolus. Bird Study, 34, 113-25. Townsend, M. G., Fletcher, M. R., Odam, E. M. & Stanley, P. I. (1981). An assessment of the secondary poisoning hazard of warfarin to Tawny Owls. J. Wildl. Manage., 45, 242-8. Worthing, C. R. & Walker, S. B. (1987). The Pesticide Manual. Lavenham Press, The British Crop Protection Council, Thornton Heath.
Rodenticides in British Barn O w l s
I. Newton, I. Wyllie & P. Freestone Institute of Terrestrial Ecology, Monks Wood Experimental Station,
Abbots Ripton, Huntingdon, Cambs PE17 2LS, UK (Received 14 December 1989; revised version received 4 July 1990; accepted 6 July 1990)
ABSTRACT Out of 145 Barn Owls found dead through accidents (66%), starvation (32%), shooting ( 2 % ) and poisoning ( < 1% ), 10% contained residues of rodenticides, difenacoum or brodifacoum, in their livers. Difenacoum was in the range 0"005q9"106 I~gg- l fresh weight, and brodifacoum was in the range 0"019-0"515 I~gg - ~. Minimum levels of detection were about 0"005 #g g - 1.for both chemicals. Mice fed for 1 day on food containing difenacoum and brodifacoum died after 2-11 days. Within these mice residues were present at greater concentration in the liver than in the rest of the carcass. The mean mass of residue in a whole 35g mouse was estimated at lO'17t~g (range 4"7320"65 l~g) for difenacoum and 15"36 pg (range 8.07-26.55)for brodifacoum. Such poisoned mice were fed to Barn Owls for successive periods of 1, 3 and 6 days. All six owls fed on difenacoum-dosed mice survived all three treatments, in which up to an estimated lO1.7#g of difenacoum was consumed, and the coagulation times of their blood returned to near normal in less than 5 23 days. Four of the six owls fed on brodifacoum-dosed mice died 6-17 days after the 1-day treatment, but the survivors also survived the 3-day and 6-day treatments. Those that died had each eaten 3 mice, with a combined weight of about 105 g and a total brodifacoum content of about 46.07#g, which was equivalent to a dose of O'15Oqg.182 mg kg-1 of owl body weight. After death these owls had 0"63-1"25 i~gg-x of brodifacoum in their livers. Blood from the survivors would not coagulate at 9 days post-treatment, but did so at 16 days in one bird and between 38 and 78 days in the other. It is concluded that: (1) Barn Owls in Britain are now widely exposed to second-generation rodenticides; (2) not all owls exposed to these chemicals 101 Environ. Pollut. 0269-7491/90/$03.50 © 1990 Elsevier Science Publishers Ltd, England. Printed in Great Britain
102
I. Newton, L Wyllie, P. Freestone are likely to receive a lethal dose; (3) brodifacoum is more toxic to owls than difenacoum; and (4) while there is yet no evidence that rodenticides have had an), appreciable effect on Barn Owl populations in Britain,further monitoring of residue levels and population trends is desirable.
INTRODUCTION This paper has two aims. The first is to report the incidence of two rodenticides, difenacoum and brodifacoum, in the bodies of Barn Owls Tyto alba found dead in various parts of Britain. The second is to report the effects of feeding captive Barn Owls on mice killed by these two rodenticides. The chemicals concerned are two of the second-generation rodenticides, developed as replacements to warfarin and other first generation rodenticides, to which rodents in some regions have become resistant. Like warfarin, both chemicals are coumarin derivatives (based on the 4-hydroxy coumarin moeity) and they act as anti-coagulants. They are more toxic and persistent than warfarin, however, and have been reported to cause secondary poisoning in various rodent-predators, including owls (Mendenhall & Pank, 1980, Hegdal & Colvin, 1988). The work described here was intended to help in assessing the impact (if any) of these new rodenticides on Barn Owls in Britain, which are thought to have declined in numbers in recent decades (Bunn et al., 1982; Shawyer, 1987). The chemical name of difenacoum is given as 3-(3-biphenyl-4-yl1,2,3,4-tetrahydro-l-naphthyl)-4-hydroxycoumarin, and of brodifacoum as 3- [3-(4'-bromobiphenyl-4-yl)- 1,2,3,4-tetrahydro- 1-naphthyl]-4-hydroxycoumarin (Worthing & Walker, 1987). Difenacoum has been available in Britain since 1975, and has been used both indoors and outdoors. Brodifacoum has been available since 1982, but apart from some experimental outdoor trials, its use has been restricted to within buildings by professional operators. Both chemicals have been commonly used around farmsteads and other sites, where rodents are a problem. As farm buildings provide some of the most frequently used roosts and nest-sites for Barn Owls in Britain, these owls would be expected to be at risk, as they eat affected rats and mice. Brown et al. (1988) listed five Barn Owls among various rodenticide victims examined during 1973-86. Likewise, Shawyer (1987) listed four Barn Owls found dead or dying, associated with local difenacoum use, and four others following brodifacoum use; all showed symptoms of haemorrhaging. He questioned whether these chemicals may have contributed to a decline in Barn Owl numbers in parts of Britain. In a more extreme situation, Duckett (1984) attributed the collapse of a Barn Owl population in a Malayan oil palm plantation to use of
Rodenticides in British barn owls
103
brodifacoum and coumachlor against rats which formed the main prey. Meanwhile, the high toxicity of brodifacoum to wild or captive owls of various species has been confirmed by Mendenhall & Pank (1980), Merson et al. (1984) and Hegdal and Colvin (1988). To our knowledge, however, there has been no previous survey of the incidence of rodenticides in wild Barn Owls and no measurement of tissue levels in Barn Owls known to have died of rodenticide poisoning. METHODS
Survey Advertisements were placed in various bird journals asking for the bodies of Barn Owls found dead. All carcasses were requested, irrespective of mode of death. On receipt, the bodies were stored deep frozen until they could be examined, usually some months later. After thawing, a post-mortem was conducted by visual inspection only. Where possible, the findings were used, along with information from the sender, to determine the cause of death. Careful search was made for signs of haemorrhaging, which often follow an exposure to anticoagulants. Typical sites for rodenticide-induced haemorrhaging include the muscle on both sides of the breast bone, the leg joints, the m o u t h and the nares, as well as various internal organs. After inspection, part of the liver was removed from each bird and analysed for residues of difenacoum and brodifacoum. The method of analysis was that of Hunter (1985) modified in minor respects. Liver samples were extracted with chloroform-acetone, and the extracts were cleared of fat using Bond-Elut columns. The concentrated samples from the columns were then analysed by Varian High Pressure Liquid Chromatography (HPLC), using a M C H - 5 N micropack column and spectrofluorometer, against a standard for each c o m p o u n d . When difenacoum or brodifacoum was detected, a recovery test was done from a spiked sample of solvent to validate the identification and to correct the estimate of mass present. Recoveries from most batches were in the range 75-95%. Specimens in which no residues were detected gave blank chromatograph traces, showing that no other material with similar properties was present. In addition, however, confirmation of the chemical identity of residues in all birds in which residues were detected was obtained using an alternative H P L C column and solvents. The lower limit of detection for both compounds was around 0.01 pg, which was equivalent to 0.005-0.01 pg g - 1, depending on the weight of the sample. For calculation of average concentrations in known contaminated
104
I. Newton, L Willie, P. Freestone
mice, nil detected values were taken as 0.0075/zg g - 1. Both isomers of each c o m p o u n d are included together in the figures reported here. Concentrations are given as wet weight values, but can be converted to dry weight values by multiplying by 3.64 for liver of owls, by 4.30 for liver of poisoned mice and by 3"29 for whole carcasses of poisoned mice (see later). The values from wild owls were corrected to allow for water loss, assuming a wet to dry conversion factor in fresh liver of 3.64, the value found for captive owls.
Toxicity trials Owls were dosed with rodenticide-poisoned mice. The owls were kept individually in cages measuring 60 x 50 × 40 cm high, and were thus much less active than wild owls, and were fed on white laboratory mice, rather than on wild rodents. As owls which are fed on fresh carcasses do not need to drink, no water was provided. Some days before a feeding trial, mice were fed for 1 day (no choice) on a proprietary dosed-food mixture containing either 0.005% difenacoum or 0.002% brodifacoum. After dying, some dats later, most mice were fed to owls. The mean weight ( ___SE) of the dead mice was 34.7 _ 1.1 g. Other dead mice, from the same batches, were analysed to determine the total content of rodenticide in their bodies. The liver was analysed separately from the rest of the carcass, which was first homogenised by crushing with liquid nitrogen. All the owls were captive-bred (second or third generation in captivity), very tame, and in their first year of life. They had not previously been exposed to rodenticides, and at the start of the trials in September they were in good condition, weighing 276-370 g. They were allocated at random into two groups of six, one fed on difenacoum-treated mice and the other on brodifacoum-treated mice. Owls in each group were fed initially for 1 day on dosed mice (3 per owl)~ Those which survived this treatment were later fed for three successive days on dosed mice, and any which survived this treatment were later fed for 6 days on dosed mice. The livers of owls which died were analysed for rodenticide residues in the same way as the livers from wild owls. Before the trials, and at various stages during it, small blood samples were taken from the brachial vein in some of the owls. In each case the coagulation time was measured, while tipping the blood back and forth in a 75 m m glass capillary tube of 1 m m internal diameter. The tube contained a 1 cm metal rod (a 'flea'), and coagulation time was taken when the rod would no longer slide around within the blood. All such sampling was done at the same time of day, at 1600 h. However, in view of the risks of haemorrhaging,
Rodent&ides in British barn owls
105
through blood-sampling birds treated with anticoagulant, such sampling was kept to a minimum. No problems were encountered, as the vein closed itself soon after the needle was withdrawn. With such a crude technique little significance could be attached to small variations in coagulation times; the main purpose of the test was to check whether the blood would coagulate or not. The findings were used mainly to indicate when recovery from exposure to the rodenticide occurred, and when another series of trials could be started. RESULTS
Survey In the period 1983-89, a total of 145 Barn Owls was received from various parts of Britain for analysis. On visual inspection of the opened carcass, together with information from the sender, most were diagnosed as victims of accidents or starvation (Table 1). In only one were the typical haemorrhage symptoms of a rodenticide victim noted. On later chemical anlysis, difenacoum was detected (and confirmed) in the livers of seven individuals, brodifacoum in four, and both chemicals together in another four, making a total of 15 (10%) contaminated birds (Table 2). These birds came from various parts of Britain (Fig. 1), but about half seemed to be from warfarin-resistance areas, as depicted by Shawyer (1987). No similar chemicals, such as bromadiolone and flocoumafen, were detected. There was no obvious seasonal variation in the percentage of received carcasses that contained rodenticide and no obvious sex bias, as TABLE 1 Causes of Death Diagnosed on Post-mortem Examination of 145 Wild Barn Owls. 1983-89
Trauma (road accident) Trauma (other accident)a Starvation Shot Poisoned
Number
Number with rodenticide residues
81 14 46 3 1
6 4 4 0 1
a Includes mainly collision victims, two drowned and one electrocuted.
TABLE 2 Numbers of Wild Barn Owls Tested for Rodenticides each Year from 1983 Showing Numbers Containing Brodifacoum and Difenacoum
Year
Number tested
Difenacoum only
Brodifacoum only
Both rodenticides
% with detectable residues
1983 1984 1985 1986 1987 1988 1989= Overall
3 15 31 36 22 23 15 145
0 0 1 2 2 1 1 7
0 1 1 0 0 1 1 4
0 0 1 2 0 1 0 4
0 7 10 11 9 13 13 10
" To 31 March only.
• Brodifacoum • Difenacoum =,,Both c h e m i c a l s o None detected
¢
F
0 ~
0-~
o
~
og "o
~o
7
%
o
oo
0
0 0 O0
'o o
o
o
g~ ° o o
%L°° °. o
Fig. 1.
o o
Map showing locations of Barn Owls found dead and received for analysis.
Rodenticides in British barn owls
107
1 0 % o f m a l e s h a d residues a n d 1 1 % o f females. H o w e v e r , o n l y 7 % o f firsty e a r b i r d s h a d d e t e c t a b l e residues, c o m p a r e d w i t h 15 % o f o l d e r birds. Levels o f d i f e n a c o u m w e r e in the r a n g e 0 . 0 0 5 - 0 - 1 0 6 p g g -1 ( m e a n 0.036 # g g - 1 ) , whilst levels o f b r o d i f a c o u m w e r e in the r a n g e 0"019-0"515/~gg -~ ( m e a n 0 - 1 8 0 # g g - x ) ( T a b l e 3). T h e single b i r d d i a g n o s e d o n p o s t - m o r t e m as a r o d e n t i c i d e victim h a d 0.43/~g g - a b r o d i f a c o u m in its liver.
Toxicity trials: Mice All d o s e d m i c e died 2-11 ( m o s t l y 3-8) d a y s a f t e r dosing, w i t h n o significant difference b e t w e e n d i f e n a c o u m a n d b r o d i f a c o u m - t r e a t e d a n i m a l s (Fig. 2). C h e m i c a l a n a l y s i s o f a s u b s a m p l e s h o w e d that, f o r b o t h chemicals, the residue w a s m u c h m o r e c o n c e n t r a t e d in the liver t h a n in the rest o f the
TABLE 3 Levels (#gg ~ in Corrected Wet Weight a) of Rodenticides found in the Livers of Contaminated Barn Owls Specimen number
Date of finding
Location
Age
Sex
Levels of rodenticide Difenacoum
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
16.11.84 12.2.85 22.5.85 31.5.85 16.1.86 9.8.86 4.9.86 20.11.86 12.5.87 13.11.87 12.1.88 8.2.88 2.12.88 --.2.89 15.3.89
Hants. Notts. Kent W. Sussex Powys Wilts. Humberside Isle of Man Beds. Salop Dyfed Suffolk Norfolk Wilts. Gwynedd
Y A A A Y A A Y A A Y A A Y A
F M F F F M F M M F M F F M M
ND ND 0"052 0-024 0.010 0.051 0"019 0-106 0"024 0-005b 0-061 ND 0-034 0.012 ND
Brodifacoum 0.515 0-300 ND 0.035 ND 0.035 ND 0"070 ND ND 0-040 0-423 ND ND 0-019
A = Adult. Y = First year bird. F = Female. M = Male. a Levels were obtained on a dry weight basis and converted to wet weight values by dividing them by 3.64, the mean wet to dry ratio in fresh owl liver. This procedure was necessary because some of the specimens had dehydrated to some extent. b Close to the limit of detection. ND, not detected.
L Newton, 1. Wyllie, P. Freestone
108
P e r c e n t a g e d e a t h s of d o s e d m i c e p e r day a f t e r dosing 50
40 :~
30
.~
20
I
Brodifacoum(N:41) D i f e n a c o u m (N=113)
I 10 °
clays a f t e r d o s a g e
Fig. 2.
Time to death in mice dosed with difenacoum and brodifacoum.
carcass (Table 4). However, because the liver forms only about one-fifteenth of the weight of a mouse, the total mass of rodenticide in the rest of the carcass was substantial, especially for brodifacoum (Table 4). On the basis of our results, a 35 g mouse on the day of death contained a total of 10.17 pg of difenacoum, on average, or a total of 15.36 pg of brodifacoum.
Toxicity trials: Owls All of the owls fed on difenacoum-dosed mice survived the 1-day, 3-day and 6-day treatments (Table 5), and none showed external bleeding. After the 1day treatment, all six owls were blood-sampled 5-9 days later, and coagulation times were 'normal' (Table 6). After the 3-day treatment, the blood of one bird sampled 3 days later would not coagulate, even after 24 h, but all birds seemed to have become normal in this respect 9-23 days after treatment. With the method used, no significance could be given to the fact that, in four of the six birds, coagulation times were somewhat shorter after recovery from difenacoum than before exposure to it (Table 6). Of the six owls fed on brodifacoum-dosed mice, four (one male and three females) died 6, 10, 11 and 17 days after the 1-day treatment, with 0"631.25 pg g - 1 brodifacoum in their livers (Table 5). The weight changes in three of these birds between treatment and death were small ( + 1%, 0%, - 6%), but the fourth bird (which died 10 days post-treatment) lost 23%, declining from 332 g to 257 g. On post-mortem, all showed internal bleeding: down
109
Rodenticides in British barn owls
+l "4D
.=_
¥, oo
¢'4
+L ()
6~
<:
+l
0
0
.<
6 ~ 6 6 6 ~ 6 ~ 6 ~ .=_ =_
÷~ r'0
C)
.>_
4r
_
.=.
~r
.o E
m
()
0
()
÷l
["
z
110
I. Newton, L Wyllie, P. Freestone
TABLE 5 Results of Feeding Rodenticide-poisoned Mice to Barn Owls
Difenaeoum
Brodifaeoum
Period o f dosing (days)
Mice eaten
Number o f owls dosed
Number surviving
1 3 6 1 3 6
3 6 5-10 3 6 12
6 6 6 6 2 2
6 6 6 2 2 2
Notes: The brodifacoum levels in the livers of the 4 owls which died were calculated at 0-63, 0-77, 1.02 and 1'25/~g g-~. The gaps between the start of successive dosing periods were 11-14 and 33-35 days for difenacoum, and 77-79 and 75 days for brodifacoum.
both sides of the keel bone, and around the brain, heart, lungs or gut, so that in two birds blood filled the body cavity. The two surviving owls also survived the later 3-day and 6-day treatments with brodifacoum. Again, however, they were affected, as they showed prolonged bleeding from the mouth or feet for up to 30 days post-treatment. Blood taken from two birds 9 days after treatment would not coagulate, nor would blood taken from one of these birds (a survivor) after 38 days. However, blood from the other survivor seemed normal 16 days after treatment (Table 6). It seems, therefore, that brodifacoum is more toxic than difenacoum to Barn Owls, and in survivors, has much more prolonged effects. From knowledge of the number of mice eaten by the owls, and the average weight of rodenticide in the mouse bodies at the time, the total mass of rodenticide consumed by the owls in the 1-, 3- and 6-day treatments could be calculated (Table 7). Individuals consumed an estimated 30.5 #g, 61.0 #g and 50-9-101"7 #g difenacoum in 1, 3 and 6 days, respectively, and an estimated 46.1 #g, 92.2 #g and 184.3 #g brodifacoum. The 46.1 #g 1-day dose of brodifacoum was enough to kill four of the six owls, and was equivalent to a dose of 0"150-0"182mgkg -1 body weight for the birds concerned. However, considering the variability in residue levels in individual mice (Table 4), these figures should be regarded as no more than rough approximations. The analysis of regurgitated pellets (of fur and other undigested parts of food items) from each group of owls revealed that some of the rodenticide that was consumed was excreted unchanged in pellets. In other words, it had not passed through the gut. Five pellets collected from one group of owls
Rodenticides in British barn owls
111
TABLE 6 Coagulation Times (Minutes, Seconds) for Barn Owl Blood Samples taken on Different Dates after End of Dosing
Dijenacoum
Owl no. 1
2
3
4
5
6
Pre-treatment
2, 00
2, 30
4, 45
4, 15
2, 45
1, 45
After 5 7 8 9
1-day dose days days days days
. --2, 30
-3, 45 --
1, 30 ---
After 3 9 10 21
3-day dose days days days days
22 d a y s 23 d a y s
NC . --
.
.
.
-1, 30 --
1, 45
-.
-1, 30
. 6, 15 --
-. . --
1, 30 --
1, 45 --
Brodifacoum 7
8
1, 15
1, 45
A f t e r 1-day d o s e 9 days 16 d a y s
NC --
NC" --
38 d a y s 78 d a y s After 3-day dose 60 d a y s
-. .
5, 00 . 2, 30
. 6, 00
2, 00
--
Owl no.
Pre-treatment
24 d a y s 36 d a y s
1, 00 ----
. .
. .
9 1, 15 a ---
. .
NC 4, 30
. .
5, 30
--
10
11
12
3, 30"
4, 00
1, 15"
--
-6, O0
. . . .
--
8, 30 5, 30 . .
. . --
. . 0, 15
--
a Died before further blood samples could be taken. N C = N o c o a g u l a t i o n w i t h i n 24 h.
during the feeding trial contained a mean of 0.60 pg difenacoum (range 0-06-1-19#g), whilst six pellets from the other group contained a mean of 0.97 pg brodifacoum (range 0"08-3.86 #g). Assuming a regurgitation rate of two pellets per day, this would have reduced the total rodenticide consumption to 29.3/~g, 57"4 pg and 43.7-94.5 pg in the 1-, 3- and 6-day difenacoum trial, and to 44.1 #g, 86.4 pg and 172.8 #g in the 1-, 3- and 6-day brodifacoum trial. Again these estimates can only be rough, because of the great variation in the rodenticide content of particular pellets.
112
L Newton, L Wyllie, P. Freestone
TABLE 7 Calculation of Mass of Difenacoum and Brodifacoum Consumed by Owls in the Feeding Trials Days
Difenacoum
Brodifacoum
of dosing
1 3 6
Number
Mean mass
Total mass
Number
Mean mass
of
of
of
of
of
of
mice eaten
difenacoum per mouse (lag)
difenacoum eaten per owl (lag)
mice eaten
brodifacoum per mouse (~g)
brodifacoum eaten per owl (lag)
30"51 61"02 50-85-101"71
3 6 12
15"36 15"36 15-36
46"08 92-16 184"32
3 6 5-10
10-17 10"17 10"17
Total mass
DISCUSSION
Survey About 10% of the 145 Barn Owls examined in 1983-89 had residues of difenacoum or brodifacoum in their bodies, but this sample may not reflect the true exposure of owls to these rodenticides in the areas concerned. Almost all the carcasses were found in the open, and were victims of road traffic and other accidents, or of apparent starvation. Death from rodenticide poisoning is delayed, and is preceded by lethargy, so poisoned owls might be most likely to die in their roosts, in roof-cavities or hollow trees, where they would be unlikely to be found by the casual observer. When they are found in such situations, moreover, they are seldom fresh enough for analysis. Any such bias in the sampling is entirely conjectural, however, and as yet there is no evidence one way or the other. Recent work on mammals and birds has shown the presence of binding sites for second-generation rodenticides in the liver, which enables residues to persist for up to several months after ingestion of a single oral dose (Parmar et al., 1987; Huckle et al., 1989a, b). In effect, this means that prey animals which took sub-lethal doses could remain as sources of contamination for owls for several months; and at the same time owls in which residues were detected could have been exposed up to several months previously or on more than one occasion. It is not certain, however, to what extent rodents acquire sublethal doses of these chemicals, as a single feed would normally be lethal. As they stand, our findings imply that the exposure of Barn Owls in Britain to difenacoum and brodifacoum is now frequent and widespread.
Rodenticides in British barn owls
113
Even if brodifacoum-use were restricted to the inside of buildings, as recommended, this would not wholly protect Barn Owls, which will sometimes hunt within buildings, and at the same time could take contaminated rodents which wander outside.
Toxicity trials The feeding trials on captive owls were conducted in order to help interpret the significance of residue levels found in wild owls. As our owls had been in captivity for only 1-2 generations, and were unrelated to one another, they could be considered as genetically representative of wild Barn Owls. Brodifacoum levels in the four captive birds that died were higher than those in the eight wild birds with brodifacoum. However, three of the wild owls had liver levels of brodifacoum approaching the lowest level in captive birds (0.30-0.52#gg-1 compared with 0.63pgg-l), while the five others had levels an order of magnitude lower (0"02-0.06 #g g- 1). There could have been at least three reasons for this difference. First, not all the wild owls may have died of rodenticide poisoning, but been exposed instead to sublethal levels. This explanation would be consistent with the low likelihood of poisoned owls being found (see above), and with the fact that, for most contaminated birds, other (non-rodenticide) causes of death were diagnosed on visual inspection. An alternative explanation would be that, owing to a behavioural difference, lower levels of rodenticides are needed to kill wild owls than captive ones. Our birds, kept in cages, remained still and inactive almost all the time, moving only when disturbed or to eat the mice provided. Yet those on brodifacoum still showed prolonged external bleeding. Wild birds, which must remain highly active, are presumably much more likely to suffer severe haemorrhaging from the exertion involved. Another likely limitation of our feeding trial is that it ran overwinter, when the owls were not moulting (apart from one which started during the 6-day brodifacoum trial). If the owls had been growing feathers throughout, the opportunities for haemorrhaging may have been much greater (though this did not occur in warfarin-fed moulting Tawny Owls Strix aluco, Townsend et al., 1981). A third explanation for the lower levels in wild than captive owls centres on the possibility of breakdown of residues after death. Although the highest residue was found in one of the longest-stored birds (4 years), the possibility of long-term post-mortem breakdown needs checking. Finally, there remains the possibility that sub-lethal levels of rodenticide may predispose death from other causes, or reduce the chance of recovery from accidents. On balance, however, we feel that the first explanation, that some wild owls received only sub-lethal levels of rodenticide and died from other causes, is the most likely reason why their levels were lower than those in captive owls.
114
1. Newton, L Wyllie, P. Freestone
On this basis, no more than 2% of the owls examined would have died of rodenticide poisoning. Despite these limitations, our results clearly imply that difenacoum is less toxic to Barn Owls than brodifacoum. None of the six birds fed difenacoum died, and blood coagulation was back to 'normal' within 5-9 days. In contrast, four of the six birds fed brodifacoum died within 6-17 days after a 1-day dose, and blood from survivors would still not coagulate at 9 days post-treatment. Birds that died ate three mice, with a combined weight of about 105 g and an estimated total brodifacoum content of 46.1 pg. The fact that two other birds of similar weight survived on the same 1-day intake, together with the later 3-day and 6-day treatments, could have been due to individual differences in sensitivity or behaviour (it was not related to sex). It was also possible that, owing to limited carrying capacity of the liver (Huckle et al., 1989b) or to continuing metabolism or excretion of residues, birds on 3- and 6-day treatments did not accumulate correspondingly greater concentrations in their tissues than did birds on the 1-day treatment. A third possibility is that birds which survived the 1-day treatment may have adjusted in some way, making them better able to withstand the 3- and 6-day treatments. These findings on the differential toxicity of the two compounds are consistent with those of Mendenhall and Pank (1980), who found that five out of six Barn Owls fed brodifacoum died, compared with none out of six fed difenacoum (but three bled externally). Nonetheless, deaths from difenacoum occurred in an extreme experiment with captive Tawny Owls Strix aluco, which were fed entirely on contaminated mice until they died after 8-41 days (Anon., 1982). The high toxicity of brodifacoum was confirmed in a field trial with voles and Screech Owls Otus asio in Virginia (Hegdal & Colvin, 1988). Five owls found dead 5-37 days post-treatment contained 0.4-0.8 pg g - 1 brodifacoum in their livers, whilst four caught and killed 34-43 days later contained 0.3-0.6 pg g-1. Two others in each group had no detectable residues. The levels in the dead Screech Owls were comparable to those in our Barn Owls of 0.63-1.25 pg g-1 suggesting a similar pre-death residue accumulation in the two species. Published values for the acute LDso for brodifacoum in different domesticated birds and mammals span two orders of magnitude: from 0.27 mg kg-1 body weight in the rat to about 25 mg kg-1 in the cat (Worthing & Walker, 1987). These figures imply a huge species variation in sensitivity. The estimated lethal dose of 0.15-0-18mgkg -~ brodifacoum which killed our owls was of similar order to the lowest published figures, making the Barn Owl one of the most sensitive of the species yet investigated. Our owls obtained this dose from three mice (total biomass 105 g), which falls within the range of 1-5 'rodents' estimated by Shawyer (1987) to be necessary to kill an owl.
Rodenticides in British barn owls
115
Normally, rodenticide residues in the rodent body reach a peak soon after ingestion, and then decline progressively until the animal dies some days later. Thus, if we had fed fresh mice to our owls, soon after bait ingestion, even fewer may have been needed to provide a lethal dose. Published figures for the acute LD 5o of difenacoum range between 0.8 and 100 mg kg-1 in different bird and mammal species, but for any one species the dose was higher than for brodifacoum (Worthing & Walker, 1987). Hence, the greater toxicity of brodifacoum over difenacoum, suggested by our results, was consistent with previous laboratory findings on other species. None of our owls would have received a dose of difenacoum as large as the smallest recorded LDso for this chemical (0.17mgkg -1 in owls compared with 0-8 mg kg- ~ in mice). In conclusion, second generation rodenticides (notably brodifacoum) could present a threat to Barn Owls and other predators of rodents. The several-day period between dosing and death in rodents, their lethargic behaviour for some hours before death, and the several-month persistence of sublethal residues within the mammalian body (Rammell et al., 1984), all facilitate the contamination of owls. Moreover, rodents that are commensal with man are widely eaten by Barn Owls, albeit in small amounts. House Mouse M u s musculus remains were present in 47% of 188 pellet samples examined by Glue (1974) from various parts of Britain, while Brown Rat Rattus norvegicus remains were present in 48% of samples. In a later Irish study, House Mouse remains were found in 10 out of 15 pellet samples, while Brown Rat remains appeared in 12 (Smal, 1987). The incidence of both species in the diet is generally small, but may well increase during hard weather or when voles are scarce. Field trials have given mixed results, but, where no owl mortality was recorded, this could be attributed to the radiomarked owls not hunting near baiting sites or taking mainly non-target prey (Kaukeinen, 1982; Hegdal & Blaskiewicz, 1984). Other field studies, in which the target rodent was the main prey, have shown mortality in owls (Merson et al., 1984, Hegdal & Colvin, 1988) and even substantial population decline (Duckett, 1984). As yet, however, despite widespread exposure and occasional mortality incidents, there is no firm evidence that modern rodenticides have had any appreciable influence on the population level of Barn Owls in Britain. However, more monitoring of residues and population trends is clearly desirable.
A C K N O W L E D G E M ENTS We are grateful to Dr P. Anderson, Dr S. Dobson and Mr C. Shawyer for useful advice at the outset, to Mr J. White for animal care, to Mrs A. Asher for help with data tabulation, and to A. P. Buckle, P. J. Edwards, S. Dobson,
116
L Newton, L Wyllie, P. Freestone
P. Greig-Smith, M. R. Hadler. To M. Roberts, C. Shawyer, I. R. Taylor and an a n o n y m o u s referee for helpful comments on the manuscript, The work was commissioned and funded by the Nature Conservancy Council, and the toxicity trials were undertaken under H o m e Office Licence. REFERENCES Anon. (1982). Secondary toxicity hazard to owls from difenacoum. In Pesticide Science 1981, Agricultural Science Service Research and Development Reports, MAFF, London, HMSO, pp. 36-7. Brown, R. A., Hardy, A. R., Greig-Smith, P. N. & Edwards, P. J. (1988). Assessing the impact of rodenticides on the environment. EPPO Bulletin, 18, 283-92. Bunn, D. S., Warburton, A. B. & Wilson, R. D. S. (1982). The Barn Owl. Calton, Poyser. Duckett, J. E. (1984). Barn Owls (Tyro alba) and the 'second generation' rat-baits utilised in oil palm plantations in Peninsular Malaysia. Planter, Kuala Lumpur, 60, 3-11. Glue, D. E. (1974). Food of the Barn Owl in Britain and Ireland. Bird Study 21, 200-10. Hegdal, P. L. & Blaskiewicz, R. W. (1984). Evaluation of the potential hazard to Barn Owls of talon (brodifacoum bait) used to control Rats and House Mice. Environ. Toxicol. & Chem., 3, 167-79. Hegdal, P. L. & Colvin, B. A. (1988). Potential hazard to Eastern Screech-owls and other raptors of brodifacoum bait used for vole control in orchards. Environ. Toxicol. & Chem., 7, 245-60. Huckle, K. R., Hutson, D. H., Logan, C. J., Morrison, B. J. & Warburton, P. A. (1989a). The fate of the rodenticide flocoumafen in the rat: Retention and elimination of a single oral dose. Pestic. Sci., 25, 297-312. Huckle, K. R., Warburton, P. A., Forbes, S. & Logan, C. J. (1989b). Studies on the fate of flocoumafen in the Japanese Quail (Coturnix coturnix japonica). Xenobiotica, 19, 51-62. Hunter, K. (1985). High-performance liquid chromatographic strategies for the determination and confirmation of anticoagulant rodenticide residues in animal tissues. J. Chromatography, 321, 255-72. Kaukeinen, D. (1982). A review of the secondary poisoning hazard potential to wildlife from the use of anticoagulent rodenticides. Proc. Vertebr. Pest. Conf., 10, 141-58. Mendenhall, V. M. & Pank, L. F. (1980). Secondary poisoning of owls by anticoagulant rodenticides. Wildl. Soc. Bull., 8, 311-15. Merson, M. H., Byers, R. E. & Kaukeinen, D. E. (1984). Residues of the rodenticide brodifacoum in voles and raptors after orchard treatment. J. Wildl. Manage., 48, 212 16. Parmar, G., Bratt, H., Moore, R. & Batten, P. L. (1987). Evidence for a common binding site in vivo for the retention of anticoagulants in rat liver. Human Toxicol., 6, 431. Rammell, C. G., Hoogenboom, J. J. L., Cotter, M., Williams, J. M. & Bell, J. (1984). Brodifacoum residues in target and non-target animals following rabbit poisoning trials. N.Z.J. Exp. Agric., 12, 107-11.
Rodenticides in British barn owls
117
Shawyer, C. R. (1987). The Barn Owl in the British Isles. Its past, present and future. London, The Hawk Trust. Smal, C. M. (1987). The diet of the Barn Owl Tyto alba in southern Ireland, with reference to a recently introduced prey species--the Bank Vole Clethrionomys glareolus. Bird Study, 34, 113-25. Townsend, M. G., Fletcher, M. R., Odam, E. M. & Stanley, P. I. (1981). An assessment of the secondary poisoning hazard of warfarin to Tawny Owls. J. Wildl. Manage., 45, 242-8. Worthing, C. R. & Walker, S. B. (1987). The Pesticide Manual. Lavenham Press, The British Crop Protection Council, Thornton Heath.