Prey composition modulates exposure risk to anticoagulant rodenticides in a sentinel predator, the barn owl

Science of the Total Environment 544 (2016) 150–157

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Prey composition modulates exposure risk to anticoagulant rodenticides in a sentinel predator, the barn owl Anke Geduhn a,b,⁎, Alexandra Esther a, Detlef Schenke c, Doreen Gabriel d, Jens Jacob a a

Julius Kühn Institute, Federal Research Centre for Cultivated Plants, Institute for Plant Protection in Horticulture and Forests, Vertebrate Research, Toppheideweg 88, 48161 Münster, Germany University of Münster, Institute of Landscape Ecology, Heisenbergstrasse 2, 48149 Münster, Germany Julius Kühn Institute, Federal Research Centre for Cultivated Plants, Institute for Ecological Chemistry, Plant Analysis and Stored Product Protection, Königin-Luise-Strasse 19, 14195, Berlin, Germany d Julius Kühn Institute, Federal Research Centre for Cultivated Plants, Institute for Crop and Soil Science, Bundesallee 50, 38116 Braunschweig, Germany b c

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Anticoagulant rodenticide exposure in small mammals drives exposure risk of barn owls. • Exposure risk of barn owls depends on seasonal variation in prey composition. • Exposure risk of barn owls is highest in autumn due to contaminated prey. • Transfer of brodifacoum to barn owls is most likely via Apodemus. • Residues of the 2nd generation anticoagulant rodenticides are common in barn owls.

a r t i c l e

i n f o

Article history: Received 10 September 2015 Received in revised form 10 November 2015 Accepted 23 November 2015 Available online xxxx Editor: D. Barcelo Keywords: Small mammal Non-target Secondary poisoning Barn owl Diet Brodifacoum

a b s t r a c t Worldwide, small rodents are main prey items for many mammalian and avian predators. Some rodent species have pest potential and are managed with anticoagulant rodenticides (ARs). ARs are consumed by target and non-target small mammals and can lead to secondary exposure of predators. The development of appropriate risk mitigation strategies is important and requires detailed knowledge of AR residue pathways. From July 2011 to October 2013 we collected 2397 regurgitated barn owl (Tyto alba) pellets to analyze diet composition of owls on livestock farms in western Germany. 256 of them were fresh pellets that were collected during brodifacoum baiting. Fresh pellets and 742 liver samples of small mammals that were trapped during baiting in the same area were analyzed for residues of ARs. We calculated exposure risk of barn owls to ARs by comparing seasonal diet composition of owls with AR residue patterns in prey species. Risk was highest in autumn, when barn owls increasingly preyed on Apodemus that regularly showed AR residues, sometimes at high concentrations. The major prey species (Microtus spp.) that was consumed most frequently in summer had less potential to contribute to secondary poisoning of owls. There was no effect of AR application on prey composition. We rarely detected ARs in pellets (2 of 256 samples) but 13% of 38 prey individuals in barn owl nests were AR positive and substantiated the expected pathway. AR residues were present in 55% of 11 barn owl carcasses. Fluctuation in

⁎ Corresponding author at: Toppheideweg 88, 48161 Münster, Germany. E-mail address: [email protected] (A. Geduhn).

http://dx.doi.org/10.1016/j.scitotenv.2015.11.117 0048-9697/© 2015 Elsevier B.V. All rights reserved.

A. Geduhn et al. / Science of the Total Environment 544 (2016) 150–157

151

non-target small mammal abundance and differences in AR residue exposure patterns in prey species drives exposure risk for barn owls and probably other predators of small mammals. Exposure risk could be minimized through spatial and temporal adaption of AR applications (avoiding long baiting and non-target hot spots at farms) and through selective bait access for target animals. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Anticoagulant rodenticides (ARs) are used to control commensal pest rodents (Buckle and Smith, 2015), but they can cause non-target exposure and poisoning. Avian predators regularly carry AR residues in many regions of the world (e.g. UK: (Walker et al., 2010) France: (Lambert et al., 2007); Spain: (López-Perea et al., 2015); USA: (Murray, 2011); New-Zealand: (Eason et al., 2002)). AR poisoning has been confirmed or was highly suspected in several avian predator species (Murray, 2011; Hughes et al., 2013; Coeurdassier et al., 2014). Therefore, appropriate risk mitigation strategies are important for non-target species conservation. However, little is known about the details of the exposure pathway and factors that modulate exposure. Such knowledge is important for risk assessment and the development of appropriate risk mitigation strategies. The barn owl is a suitable model species for risk assessment of ARs because they use farm buildings (De Bruin, 1994) where ARs are regularly applied. Barn owls are increasingly exposed to ARs (Newton et al., 1997; Walker et al., 2010) and numbers are declining in the long-term in Germany (Bundesamt für Naturschutz, 2009). AR bait use in Germany requires covered application (Umweltbundesamt, 2014) to minimize primary poisoning of non-targets. Secondary poisoning can occur via target and non-target small mammals of rodent control programs. Main target species of biocidal AR application are Rattus norvegicus and Mus musculus. The latter rarely occurs in barn owls diet in Germany (Görner, 1979; Langenbach, 1982), Great Britain (Glue, 1967; Love et al., 2000), Italy (Bose and Guidali, 2001) and the United States (Smith et al., 1972) but can be common in barn owl diet in south-eastern Europe (Goutner and Alivizatos, 2003; Bontzorlos et al., 2005). Similarly, R. norvegicus rarely occurs in barn owl prey (Smith et al., 1972; Görner, 1979; Langenbach, 1982; De Bruin, 1994; Love et al., 2000; Bose and Guidali, 2001; Goutner and Alivizatos, 2003), except for local situations, where rats are hunted in considerable amounts (Bontzorlos et al., 2005; Obuch and Benda, 2009). It seems that high AR exposure of barn owls (Walker et al., 2010; López-Perea et al., 2015) is at least partly be driven by AR residues in non-target small mammals. Non-target small mammals are a considerable source of AR exposure because AR residues in non-target small mammals have been reported in the UK (Brakes and Smith, 2005; Tosh et al., 2012), Canada (Elliott et al., 2014) and Germany (Geduhn et al., 2014). Secondary exposure of predators via non-target species was discussed by Tosh et al. (2011) for red foxes in Great Britain and Ireland. However, quantitative data are scarce. The composition of barn owl diet and differences in AR exposure among small mammal species may drive the risk of AR exposure in predators. In addition, exposure risk could vary because of seasonal differences in prey abundance and seasonal variation in the use of ARs for rodent management (Shore et al., 2003). Huson and Rennison (1981) found increasing rat infestations on agricultural premises in England from late summer to winter. They suggest that rats occur on farms when food availability decreases after harvest and found farmers controlling rats mainly in winter. This enhances risk of AR-exposure of predators during this period. For developing optimal risk mitigation strategies, detailed information about exposure pathways from bait to predators is required. Therefore, we considered a) species composition of barn owl diet (targets/ non-targets) from a period of 28 months and b) combined this and exposure patterns of non-target small mammals (based on Geduhn et al.

(2014)) to estimate exposure risk for barn owls. We additionally analyzed the influence of c) seasonal variation in barn owl diet. Secondly, the expected exposure pathway was assessed. We screened d) pellets of barn owls and e) prey caught by barn owls and carcasses of barn owls for AR residues. 2. Material and methods 2.1. Samples and study area The risk assessment for barn owl exposure to ARs was based on the AR residues in small mammals that were trapped during baiting campaigns at livestock farms in the Münsterland region (52°N, 8°E) in western Germany and the diet composition of barn owls in the same area. The expected exposure pathway was monitored by the analysis of AR residues in barn owl pellets and liver samples from prey individuals that were hunted by the owls at the same farms. The study area where the barn owl pellets and small mammals were originated from is a mosaic of farmland (about 60%) interspersed by small forest sections (about 15%). Nesting boxes for barn owls were available at all 9 investigated livestock farms. We used brodifacoum (BR) bait (Ratron® Brodifacoum Flocken 0.05 g/kg BR, frunol delicia® GmbH) to control R. norvegicus in October/November at 6 (2011) and 9 farms (2012) and in February/March at 7 (2012) and 8 farms (2013). Baiting campaigns lasted three weeks following label instructions. Further farm details and anticoagulant rodenticide use is described in Geduhn et al. (2014). From July 2011 until October 2013 we collected barn owl regurgitated pellets from rest and nesting sites at livestock farms once a month (monitoring pellets n = 2141) and every third day during baiting campaigns (fresh pellets n = 256). The barn owl diet was assessed by both, monitoring and fresh pellets and fresh pellets were analyzed for residues of ARs. During and one week after the baiting campaigns we trapped 742 small mammals on farms up to 100 m away from bait points (Geduhn et al., 2014). Furthermore, we collected prey carcasses (n = 38 small mammals) from barn owl nest boxes, which were checked every third day during the baiting campaign in February/March 2012. Whole liver samples of all small mammals were analyzed for AR residues. In addition, barn owl carcasses at regional scale (three German federal states: North Rhine-Westphalia, Lower Saxony and BadenWuerttemberg) were analyzed for AR residues. 11 liver samples were obtained from a veterinarian practitioner, two veterinary institutes and by us. Barn owls were found dead or were euthanized shortly after admission of moribund individuals to a veterinarian. 2.2. Barn owl diet analysis After initially removing all barn owl pellets present, we collected monitoring pellets once a month (sampling occasion) and dried them for at least three hours at 100 °C. We soaked pellets in tap water, disintegrated them with a strong jet of water and collected all solid components on a mesh. Solids were placed in a plastic bowl and a finer jet of water was used to separate bones and teeth from hair. Hair was decanted and bones and teeth were collected and dried at 60 °C. We identified prey species by cranium and teeth characteristics (Jenrich et al., 2010) and recorded the minimum number of prey individuals by counting upper and lower jaws. Mean body weight of prey

152

A. Geduhn et al. / Science of the Total Environment 544 (2016) 150–157

species (of trapped small mammals or from literature; Table 1) was used to calculate the prey species' contribution to total prey biomass. We collected fresh pellets every third day for five weeks (one week before, three weeks during and one week after baiting) and stored pellets at − 20 °C until analysis. Pellets were defrosted and we separated craniums and teeth from fur manually to identify species and pooled data with monitoring pellets (described above). Pellet remains were used for rodenticide residue analysis (see 2.3). We selected farms where at least 5 fresh pellets were available per period and at least 10 prey individuals were identified to species to compare barn owls diet composition before and during/after the baiting campaign. We used paired sampled t-tests to test for differences in prey composition before and during/after the baiting campaign. To model seasonal variation in the composition of barn owl diet we used a general additive mixed effect model with the proportion of each taxon in prey occurrence per farm and month as depending variable. We fitted a smother term for month (January to December) for the five most relevant taxa (Microtus, Myodes glareolus, Apodemus, Crocidura russula, Sorex) and included year (2011 to 2013) and taxon as main and interaction effects (i.e. fixed effects). To account for the repeated sampling, we further included sampling occasion (1–104) nested with farm (1–6) as random effect in the model. We compared fits of models containing different sets of fixed effects based on the Akaike Information Criterion. For this analysis, data of 6 farms were available with at least nine observations from July 2011–October 2013 and a minimum observation of 5 pellets and 10 prey individuals per month. All statistical analyses were conducted with R (RCoreTeam, 2013) and the package mgcv (Wood, 2007). Level for significance was p b 0.05.

2.3. Chemical AR residue analysis Whole liver samples of small mammals and 1.5 g subsamples of barn owl livers were used to analyze samples for five second generation ARs (SGARs): brodifacoum, bromadiolone, difenacoum, difethialone, flocoumafen and three first generation ARs (FGAR): chlorophacinone, coumatetralyl, warfarin. Liquid chromatograph UltiMate 3000 RS (Dionex) coupled with the mass spectrometer QTRAP 5500 (AB SCIEX) used in electrospray ionization mode provided concentration values for AR residues. Trapping, sample purification and HPLC settings are described in Geduhn et al. (2014). We corrected residue concentrations by recovery rates of ARs in blank turkey liver. Remains of fresh barn owl pellets were also analyzed for AR residues. We used the same residue analysis method for regurgitated barn owl pellets as for liver samples (Geduhn et al., 2014) but froze pellets in liquid nitrogen for homogenization with a stainless steel mortar. Recovery rates for barn owl pellets were determined in fresh barn owl pellets that were collected in summer and were spiked with all eight substances (recovery rate, relative standard deviation): brodifacoum 83 (8), bromadiolone 76 (5), chlorophacinone 133 (14), coumatetralyl 73 (7), difenacoum 132 (10), difethialone 52 (10), flocoumafen 70 (8), warfarin 116 (6). 2.4. Risk assessment Risk assessment was based on the % occurrence of prey in the barn owl diet based on pellets that were collected during months of the baiting campaigns (monitoring and fresh pellets) and the BR amount in the liver of the five most commonly trapped small mammal taxa.

Table 1 Prey of barn owls in western Germany. Data are based on barn owl pellets (2397 samples) sampled July from 2011 to October 2013 at nesting and resting sites at livestock farms in the Münsterland region.

Voles Arvicola spp. Microtus agrestis Microtus arvalis Microtus subterraneus Microtus spp. Myodes glareolus Mice Apodemus flavicollis Apodemus sylvaticus Apodemus spp. Micromys minutus Mus musculus Rattus norvegicus Shrews Crocidura russula Neomys fodiens Sorex araneus/coronatus Sorex minutus Other Aves Talpa europaea Small mammal a b c d e f g h i j k

No.a

%b

[g]c

% massd

Max. %/farm/monthe

Max. ind./pelletf

Av. ind./pelletg

142 392 3299 189 79 488

1.7 4.8 40.5 2.3 1.0 6.0

95.5h 21.8 17.6 15.0h 18.0i 19.3

9.6 6.1 41.3 2.0 1.0 6.7

17.5 16.7 83.3 9.8 10.9 30.7

3 3 9 3 3 4

1.07 1.12 2.10 1.16 1.07 1.28

280 866 74 92 54 26

3.4 10.6 0.9 1.1 0.7 0.3

28.4 20.8 24.6j 6.9 19.5 66.5k

5.7 12.8 1.3 0.5 0.7 1.2

25.3 68.6 9.1 12.5 8.7 12.1

3 4 4 5 3 2

1.10 1.32 1.10 1.15 1.15 1.08

213 30 1492 358

2.6 0.4 18.3 4.4

11.8 12.7h 7.7 3.7

1.8 0.3 8.2 0.9

30.3 9.1 72.2 26.1

6 2 8 6

1.48 1.07 1.84 1.34

26 1 44

0.3 b0.1 0.5

– 61.4 –

b0.1

5.9 2.3

1 1 2

1.00 1.00 1.10

Number of identified prey items. The percentage of a taxon in diet. The average biomass of a species based on body weight derived from small mammals that were trapped for Geduhn et al. (2014) or from the literature (h/k). The percentage of a taxon by biomass. The maximal percentage of a taxon on a specific farm and month. The maximal number of prey individuals of a taxon within one pellet. The average number of individuals per pellet (calculated for pellets with at least one prey item of this species). Braun and Dieterlen (2005). Mean adult body weight of M. agrestis, M. arvalis and M. subterraneus. Mean adult body weight of A. flavicollis and A. sylvaticus. Morris (1979).

A. Geduhn et al. / Science of the Total Environment 544 (2016) 150–157

We multiplied mean BR occurrence with BR liver content (μg/liver) of each small mammal taxon to obtain mean μg BR (BR dose) potentially transferred to owls. Pellets from barn owl nests investigated in the baiting period were used to calculate mean occurrence of each prey taxon per farm. Exposure risk for barn owls to BR was calculated by multiplying BR doses by relevant small mammal taxa and occurrence (ratio) of these taxa in barn owl diet. BR residue occurrence and concentrations were not significantly different between season (Geduhn et al., 2014), whereas barn owl diet was (see results “Barn owl diet”). Therefore, we separated data of Oct./Nov. and Feb./Mar. for barn owl diet but not for residues in small mammals and calculated the risk for both seasons separately.

3. Results 3.1. Barn owl diet A total of 8145 prey individuals were identified in 2379 barn owl pellets (mean 3.4, maximum 12). Microtus species were the predominant prey contributing to 48.6% of prey items and 50.4% of prey biomass (Table 1). Apodemus species and Sorex shrews occurred regularly and M. musculus and R. norvegicus rarely in diet. The percentages of occurrence of the main taxa in barn owl diet did not differ between before (110 pellets and 412 prey items) and during/ after (258 pellets and 771 prey items) baiting (farms: N = 12 (3 autumn 2011, 3 winter 2012, 3 autumn 2012 and 3 winter 2012); Apodemus spp.: t = − 1.4, p = 0.192; Microtus spp.: t = 1.6, p = 0.135; M. glareolus: t = 0.34, p = 0.74; C. russula: t = 0.58, p = 0.577; Sorex spp.: t = − 0.50, p = 0.626; Fig. 1). M. musculus and R. norvegicus occurred sporadically before and during/after baiting. The mean percentage of small mammal taxa in the barn owl diet varied among months (Fig. 2a/b). The model explained 69.3% of deviance in monthly prey occurrence. Microtus was the predominant taxon in most months with annual peaks in summer (August–September) and winter (January, Fig. 2b). There was a significant effect of month on the occurrence of Microtus, Apodemus and Sorex in prey (Table 2). Remarkably, Apodemus occurrence was high in April to June and in November (Fig. 2) and Sorex occurrence was high in March and April (Fig. 2). Commensal rodents rarely occurred in the diet. R. norvegicus were present only from November 2011 to January 2012, from October 2012 to February 2013 and in June and August to October 2013 reaching a maximum of 12.1% on a farm in January 2013. We found M. musculus at low numbers but almost all year round (0–1.7%). C. russula also occurred almost continuously but at 0–9.1%; maximum 30.3% (Fig. 2).

153

3.2. Risk assessment: based on barn owl diet and rodenticide residues in small mammals BR residue occurrence and concentrations in trapped small mammals based on Geduhn et al. (2014) were used to calculate the average dose of BR provided by an individual of a taxon. (Table 3). BR was present in 66% of C. russula (Table 3) mainly at low residue concentrations. In Microtus species, which were the main barn owl prey, BR residues were rare (7%). BR residue occurrence and mean concentrations were similar in Apodemus and M. glareolus (Geduhn et al., 2014) but Apodemus occurred more often in the barn owl diet, especially in autumn (see Section 3.1). Also Sorex was regularly contaminated with BR (28%) and made up 29.9% of the owl diet in Feb./Mar. Nevertheless, exposure risk of owls due to Sorex was low because mean BR amount per liver was low. Maximum BR concentrations were highest in Apodemus (53.663 μg/g) and lowest in Sorex (1.320 μg/g) (Table 3). However, median BR residue concentrations were highest in Microtus and Sorex. 3.3. AR-residues in barn owl pellets and carcasses of prey and barn owls We identified 270 Microtus spp., 137 Apodemus spp., 222 Sorex spp., 56 M. glareolus, 12 M. musculus, 11 C. russula, 3 Arvicola spp., Micromys minutus and Neomys fodiens, 2 R. norvegicus and 5 birds in 256 fresh pellets. AR residues of three compounds were detected in 1.6% (4) of owls pellets that were collected before (24) and during/after (232) baiting campaigns. BR residues were present in two pellets. One contained a bird skull (no mammals) and was collected during the baiting campaign (0.147 μg/g). The other pellet contained two Apodemus, one M. glareolus and one Sorex and was sampled before the baiting campaign started (0.011 μg/g). One pellet contained 0.007 μg/g coumatetralyl (2 Apodemus, 1 M. glareolus) and another 0.010 μg/g flocoumafen. Both were collected during the baiting period. BR residues were detected in 13% (5 of 38) liver samples of small mammals that were caught by barn owls during the baiting campaign in March 2012. BR-positive small mammals occurred at 3 of 4 farms sampled. Three Apodemus (21), one Microtus (8) and one R. norvegicus (1) showed BR residues. BR concentrations were 0.021–0.122 μg/g. There were no residues in M. glareolus (3), M. musculus (1) and Sorex (4). AR residues were present in 55% of 11 liver samples of barn owl carcasses. One sample contained 3 active substances (BR, bromadiolone and flocoumafen) while the five other samples contained one active substance. We identified residues of SGARs in 5 of 11 samples, 3 carrying SGAR residues N0.2 μg/g. BR and flocoumafen residues were most prevalent (each 3 samples). Coumatetralyl was the only FGAR present (1 sample). Two barn owls were found dead at two farms where baiting

Fig. 1. Barn owl diet composition before and during/after baiting campaigns. Brodifacoum was applied on livestock farms in the Münsterland region in western Germany. Data are mean occurrence ± standard deviation.

154

A. Geduhn et al. / Science of the Total Environment 544 (2016) 150–157

Fig. 2. Seasonal variation in barn owl diet. (a): Monthly abundance of prey species in barn owl diet in western Germany based on barn owl pellets sampled July 2011 to October 2013 at nesting and resting sites at livestock farms in the Münsterland region. Data are mean percentage ± standard deviation of farms. Arrows mark bait application. (b): Modeled seasonal variation in barn owl diet. Dotted lines represent the 95% confidence interval.

with BR was common practice carried residues of 0.019 and 0.233 μg/g BR respectively. 4. Discussion Prey composition and differences in AR residue occurrence and concentration in prey taxa drive AR exposure risk in barn owls and possibly other predator species of small mammals. We quantified the risk with replicated field data and followed the exposure pathway from bait to predators. Our findings create a basis for the development of appropriate risk mitigation strategies to minimize the effect of biocidal AR application on non-target species and therefore contribute to non-target conservation. 4.1. Barn owl diet In our study, barn owl diet was dominated by non-target small mammals (Table 1), mainly Microtus (48.6% of individuals; 50.4% of biomass), followed by Sorex (22.7% of individuals) and Apodemus (19.8% of biomass). Microtus mainly consisted of Microtus arvalis, a common

species of central European grasslands (Jacob and Tkadlec, 2010). Microtus species are common in barn owl diet in several parts of Europe (UK: (Glue, 1967; Love et al., 2000), Italy: (Bose and Guidali, 2001), Netherlands: (De Bruin, 1994)) contributing to 24–45% of prey items. The low body weight of Microtus in our study was driven by juvenile individuals that were caught in autumn. However, we used this value because barn owls are likely to catch juveniles as well as adults. At a mean body weight of about 28 g (19-37 g range) of adult individuals (Jacob, 2003), Microtus would comprise 53% of biomass of the barn owl diet. Barn owls seemed to hunt at least occasionally in the direct surrounding of farms because the human associated C. russula occurred occasionally in barn owl diet (mean 2.6%) but sometimes quite regularly (max. per farm 30% in November 2011 and up to 6 individuals per pellet). The occurrence of target rodent species of baiting campaigns (R. norvegicus and M. musculus) in barn owl diet was low. Therefore, AR transfer from prey to barn owls was mainly driven by non-target prey taxa. The composition of barn owl diet was affected by month. This was present in Microtus, Apodemus and Sorex. Microtus species dominated barn owl diet especially in summer, whereas Apodemus mainly occurred

A. Geduhn et al. / Science of the Total Environment 544 (2016) 150–157

intensify rodenticide use ((Birks, 1998) and personal communication with farmers in the Muensterland region). This period overlapped with the increased occurrence of Apodemus in the barn owl diet (Fig. 2), increasing AR exposure risk for barn owls mainly in autumn. Furthermore, M. glareolus occurrence in the diet was higher in November/December by trend, although at 13% only. Despite high occurrence of Microtus in barn owl diet and high mean BR concentration residues in Microtus we rarely found residues in this taxon suggesting reduced overall risk linked to Microtus. Microtus were present in high percentages in the barn owls' diet the whole year, but especially high peaks were found in August/September. Risk for AR poisoning in barn owls via shrews was low as C. russula rarely occurred in barn owl diet and residue concentrations in Sorex were consistently low. In October/November there was a higher risk for barn owls via C. russula when owls ingested several individuals at once, which can occur as indicated by our data. In spring, Sorex occurred often in barn owl diet and regularly carried BR residues but at low concentrations. Therefore, Sorex and Microtus occurrence in barn owl diet lead to a lowered exposure risk for barn owls in spring and summer. Median residue concentrations were especially high in Microtus and Sorex species. However, using median residue concentrations would weaken the effect of those individuals that carried high BR residues because the majority of animals were with low residues. Mainly individuals that carried those high residues cause AR exposure for barn owls. For example, Sorex is a taxa that carried residues that often were around 0.25 μg/g liver wet weight, but never had residues above 1.5 μg/g. In contrast many Apodemus individuals carried residues below 0.1 μg/g but some individuals carried residues that were higher than 10 μg/g. These individuals are important when calculating poisoning risk for barn owls. Therefore, we used mean residue concentrations to assess exposure and risk. We used liver samples of small mammals for AR residue analysis. However, other tissues of the carcasses also contain residues. Sage et al., 2008 found about 25% of whole body bromadiolone residues in the liver of water voles. Assuming similar residue distribution of BR in the small mammals considered in this study, the absolute risk for barn owls would be 4 times higher than stated. However, relative risk among small mammal taxa would be similar.

Table 2 Seasonal variation of barn owl diet: statistical modeling results. Smoother

Estimated degrees of freedom

S (month) ∗ Apodemus spp. S (month) ∗ Crocidura russula S (month) ∗ Myodes glareolus S (month) ∗ Microtus spp. S (month) ∗ Sorex spp.

5.888 1.000 1.885 6.844 6.562

Fixed effects

Estimate

Intercept (2011/Apodemus spp.) 2012 2013 Crocidura russula Myodes glareolus Microtus spp. Sorex spp. 2012 ∗ Crocidura russula 2013 ∗ Crocidura russula 2012 ∗ Myodes glareolus 2013 ∗ Myodes glareolus 2012 ∗ Microtus spp. 2013 ∗ Microtus spp. 2012 ∗ Sorex spp. 2013 ∗ Sorex spp.

0.240 −0.031 −0.170 −0.215 −0.148 0.171 −0.069 0.032 0.165 0.018 0.111 0.067 0.268 0.057 0.296

F

p

5.188 0.002 2.876 11.045 6.980

b0.001 0.965 0.061 b0.001 b0.001

Std. error

p b0.001 0.344 b0.001 b0.001 b0.001 b0.001 0.068 0.483 b0.001 0.692 0.027 0.139 b0.001 0.217 b0.001

0.026 0.032 0.036 0.037 0.037 0.037 0.037 0.045 0.050 0.045 0.050 0.045 0.051 0.046 0.051

155

in spring and autumn and Sorex in spring. We found no effect of the baiting campaign on prey composition. This may indicate that there was no effect of baiting on non-target population dynamics or that poisoned small mammals are not preferentially hunted by barn owls. 4.2. Risk assessment: based on barn owl diet and rodenticide residues in small mammals Our results demonstrate that AR exposure risk of barn owls varied seasonally depending on prey composition. The highest exposure risk for barn owls was in October/November, when they increasingly preyed on Apodemus which regularly carried residues at occasionally high concentrations (Table 3). Furthermore, Apodemus was trapped most often during and after baiting (Geduhn et al., 2014). In February/March Apodemus was less common in barn owl diet suggesting a reduced risk to barn owls. Nevertheless, exposure risk to barn owls was still highest via Apodemus because of higher occurrence of Apodemus in barn owl diet compared to Myodes. In autumn and winter food availability for commensal rodents is low in the field but still high on farms. As a consequence, rat infestations post-harvest are high and farmers

4.3. AR-residues in barn owl pellets and carcasses of prey and barn owls Less than 1% of pellets analyzed contained BR residues. This suggests that barn owls hunt away from baited areas. However, several aspects point towards owls occasionally hunting in the farm environment: first: commensal small mammals were sometimes present in owl diet.

Table 3 Risk assessment for brodifacoum (BR) exposure of barn owls. Risk assessment is based on the percentage of occurrence (% BR), and the amount of BR (mean μg/liver) in BR positive small mammals during and after baiting campaigns in the Münsterland region of western Germany (Geduhn et al., 2014) and the diet of barn owls (% owl diet) in the same area. Taxa

BR residues of pos. ind. [μg/g]a

BR occurrence % BR

Mean

Median

Max.

Mean μg/liver

BR dose

130

7

9

4.062

0.414

13.775

3.330

0.233

Myodes

168

26

43

3.290

0.062

34.793

4.671

1.214

Apodemus

307

21

65

4.676

0.102

53.663

3.958

0.831

Crocidura

38

66

25

0.989

0.091

7.391

0.698

0.461

Sorex

89

28

25

0.369

0.283

1.321

0.160

0.045

a

N (pos)

Riskb

Barn owl diet

Microtus

N

d

BR amount

c

e

Season

% owl diet

Oct./Nov. Feb./Mar. Oct./Nov. Feb./Mar. Oct./Nov. Feb./Mar. Oct./Nov. Feb./Mar. Oct./Nov. Feb./Mar.

36.7 41.6 13.4 6.6 29.5 14.1 3.8 2.1 11.1 29.9

f

0.086 0.097 0.163 0.080 0.245 0.117 0.018 0.010 0.005 0.013

Of individuals with residues. In contrast to Geduhn et al. (2014) residue concentrations were corrected for recovery rates. Barn owl risk (risk) was calculated by multiplying BR doses with % occurrence of the prey taxon in the owl diet. Number of individuals analyzed. d Number of individuals with BR residues. e The percentage of occurrence (% BR) was multiplied with the mean amount of BR per liver (mean μg/liver) to calculate the mean BR dose contained per liver of a prey individual (BR dose) of a specific taxon. f Based on pellets that were collected during the months baiting was conducted (Oct./Nov. and Feb./Mar.). b c

156

A. Geduhn et al. / Science of the Total Environment 544 (2016) 150–157

Second: 13% of small mammals from barn owl nest boxes and third: 55% of barn owl carcasses collected in our study carried AR residues. This raises the question whether analytical methods were entirely appropriate to detect ARs in owl pellets. On average 28% of the ingested rodenticide is regurgitated by barn owls via pellets in feeding trails (Gray et al., 1994; Newton et al., 1994). However, in pellets collected in 2–8 day intervals from sites where many nearby farmers used rodenticides no residues were found (Eadsforth et al., 1996), but Elliott et al. (2014) identified residues of difethialone and chlorophacinone in 9 of 9 barn owl pellets that contained remains of R. norvegicus. In feeding studies most flocoumafen was excreted within two days after ingestion (Eadsforth et al., 1991; Newton et al., 1994) and probably spread across 3–4 pellets as barn owls regurgitate on average 1.7 pellets per night (Marti, 1973). Therefore, the concentration in one pellet could be close to the limit of detection. Schenke et al. (2013) found that brodifacoum solved in acetonitril was degenerated considerably in one week in light and at room temperature. The low starting concentration of BR in a given pellet combined with rapid degradation of BR at ambient conditions may explain the lower than expected occurrence of BR residues in owl pellets. AR residues were found in 13% of small mammals collected from barn owl nests during baiting campaigns. This demonstrates that barn owls sometimes catch AR-exposed small mammals. However, residue concentrations in such prey individuals were always low (max. 0.071 μg/g). Low concentrations of ARs were also found in many small mammals trapped in the study but low concentrations in prey from owl nests could also result from degradation of AR, because sites were visited every third day only compared to sampling small mammals by trapping within 24 h. We found residues in 55% of 11 barn owls. AR occurrence was similar to that in red foxes (60%) in Germany (Geduhn et al., 2015) but lower than in most recent studies of avian predators (e.g. López-Perea et al. (2015); Murray (2011) and Hughes et al. (2013) ranging from 63 to 86%). This may due to differences in the analytical method (UV-, Fluorescence- or MS-detection) but also to differences in the conformation criteria for low level residues (e.g. Geduhn et al. (2014); LópezPerea et al. (2015); Murray (2011) and Hughes et al. (2013). However, the abstinence from AR use in agriculture during recent years in Germany may have contributed to relatively low residue levels also. AR residue occurrence in barn owls in our study area substantiates the expected pathway via non-target small mammals because barn owls almost only prey on such species. Hardly any residues of FGARs could be detected, probably because of their short half-life in liver compared to SGARs and a less common usage of biocidal FGARs because of their ineffectiveness against commensal rodents that are genetically resistant to FGARs (Lund, 1964; Pelz, 2007). In our study area farmers mainly used BR bait (Geduhn et al., 2014) and this substance and flocoumafen we found most often in carcasses of barn owls. This confirms the result of a recent study in Spain where mostly BR occurred in non-target species (López-Perea et al., 2015). Newton et al. (1990) detected BR residue concentrations of 0.63 to 1.25 μg/g in livers of barn owls that had died after consuming three poisoned mice. Based on pathologic examination AR poisoning was confirmed at b 0.1 μg/g SGAR (Stone et al., 1999; Thomas et al., 2011). Modeling indicated an 11% mortality risk in barn owls at 0.1 μg/g SGAR liver concentration, 22% at 0.2 μg/g and 54% at 0.7 μg/g (Thomas et al., 2011). We found residues above 0.2 μg/g SGAR in 3 of 11 liver samples, suggesting at least some negative biological effects. 5. Conclusion We conclude that non-target small mammals drive AR exposure in barn owls because we regularly found residues in small mammals associated to bait application on farms. Barn owls preyed on these taxa and prey items included exposed individuals. Furthermore, AR residues occurred in 6 of 11 barn owl carcasses. Nevertheless, residue analysis of

pellets was not the most appropriate method to monitor exposure in barn owls as almost no residues could be detected in pellets. Based on the identification of AR exposure risk regarding prey taxa and season, risk mitigation can be optimized through spatial and temporal adaption of AR applications. The baiting period should be restricted to the necessary minimum especially in autumn when risk for barn owls is highest. Restricting bait access to targets will minimize direct and indirect nontarget species exposure. The size difference between R. norvegicus and non-target small mammals (mice, voles and shrews) could be used to achieve such a selective bait access. Furthermore, structures on farms those are associated with an increased occurrence of non-target small mammals but not with target species should be identified and excluded from baiting. Further research is needed to develop these potential options for risk mitigation. Acknowledgments We thank all farmers for providing access to their farms and E. Hegemann, F. Brandes, H. Meinig, and the CVUA Freiburg for barn owl liver acquisition and sampling. Further thanks go to R. Koç, M. Hoffmann and I. Stachewicz-Voigt for residue analysis. This study was funded by the German Federal Environment Agency within the Environment Research Plan of the German Federal Ministry for the Environment, Nature Conservation and Nuclear Safety (grant number 3710 63 401). References Birks, J., 1998. Secondary rodenticide poisoning risk arising from winter farmyard use by the European polecat (Mustela putorius). Biol. Conserv. 85, 233–240. Bontzorlos, V.A., Peris, S.J., Vlacho, C.G., Bakaloudis, D.E., 2005. The diet of barn owl in the agricultural landscapes of central Greece. Folia Zool. 54, 99–110. Bose, M., Guidali, F., 2001. Seasonal and geographic differences in the diet of the barn owl in an agro-ecosystem in northern Italy. Journal of Raptor Research 35, 240–246. Brakes, C.R., Smith, R.H., 2005. Exposure of non-target small mammals to rodenticides: short-term effects, recovery and implications for secondary poisoning. J. Appl. Ecol. 42, 118–128. Braun, M., Dieterlen, F., 2005. Die Säugetiere Baden-Württembergs. Ulmer Verlag, Stuttgart. Buckle, A.P., Smith, R.H. (Eds.), 2015. Rodent Pests and Their Control, Second edition CAB International, Wallingford, Oxfordshire. Bundesamt für Naturschutz, 2009. Rote Liste gefährdeter Tiere, Pflanzen und Pilze Deutschlands. Landwirtschaftsverlag, Münster. Coeurdassier, M., Riols, R., Decors, A., Mionnet, A., David, F., Quintaine, T., Truchetet, D., Scheifler, R., Giraudoux, P., 2014. Unintentional wildlife poisoning and proposals for sustainable management of rodents. Conserv. Biol. 28, 315–321. De Bruin, O., 1994. Population ecology and conservation of the barn owl Tyto alba in farmland habitats in Liemers and Achterhoek (The Netherlands). Ardea 82, 1–109. Eadsforth, C.V., Dutton, A.J., Harrison, E.G., 1991. A barn owl feeding study with [14C] flocoumafen-dosed mice: validation of a non-invasive method of monitoring exposure of barn owls to anticoagulant rodenticides in their prey. Pestic. Sci. 32, 105–119. Eadsforth, C.V., Gray, A., Harrison, E.G., 1996. Monitoring the exposure of barn owls to second-generation rodenticides in southern Eire. Pestic. Sci. 47, 225–233. Eason, C.T., Murphy, E.C., Wright, G.R.G., Spurr, E.B., 2002. Assessment of risks of brodifacoum to non-target birds and mammals in New Zealand. Ecotoxicology 11, 35–48. Elliott, J., Hindmarch, S., Albert, C., Emery, J., Mineau, P., Maisonneuve, F., 2014. Exposure pathways of anticoagulant rodenticides to nontarget wildlife. Environ. Monit. Assess. 186, 895–906. Geduhn, A., Esther, A., Schenke, D., Mattes, H., Jacob, J., 2014. Spatial and temporal exposure patterns in non-target small mammals during brodifacoum rat control. Sci. Total Environ. 496, 328–338. Geduhn, A., Jacob, J., Schenke, D., Keller, B., Kleinschmidt, S., Esther, A., 2015. Relation between intensity of biocide practice and residues of anticoagulant rodenticides in red foxes (Vulpes vulpes). PLoS One 10 (9), e0139191. Glue, D., 1967. Prey taken by the barn owl in England and Wales. Bird Study 14, 169–183. Görner, M., 1979. Zur Verbreitung der Kleinsäuger im Südwesten der DDR auf der Grundlage von Gewölleanalysen der Schleiereule (Tyto alba SCOP.). Zoologische Jahrbücher Systematik 106, 429–470. Goutner, V., Alivizatos, H., 2003. Diet of the barn owl (Tyto alba) and little owl (Athene noctua) in wetlands of northeastern Greece. Belg. J. Zool. 133, 15–22. Gray, A., Eadsforth, C.V., Dutton, A.J., Vaughan, J.A., 1994. Noninvasive method for monitoring the exposure of barn owls to 2nd-generation rodenticides. Pestic. Sci. 41, 339–343. Hughes, J., Sharp, E., Taylor, M.J., Melton, L., Hartley, G., 2013. Monitoring agricultural rodenticide use and secondary exposure of raptors in Scotland. Ecotoxicology 22, 974–984. Huson, L.W., Rennison, B.D., 1981. Seasonal variability of Norway rat (Rattus norvegicus) infestation of agricultural premises. J. Zool. 194, 257–260.

A. Geduhn et al. / Science of the Total Environment 544 (2016) 150–157 Jacob, J., 2003. Body weight dynamics of common voles in agro-ecosystems. Mammalia 67 (4), 559–566. Jacob, J., Tkadlec, E., 2010. Rodent outbreaks in Europe: dynamics and damage. In: Singleton, G.R., Belmain, S., Brown, P.R., Hardy, B. (Eds.), Rodent Outbreaks — Ecology and Impacts. International Rice Research Institute, Los Baños, Philippines, pp. 207–223. Jenrich, J., Löhr, P.W., Müller, F., 2010. Bildbestimmungsschlüssel für Kleinsäugerschädel aus Gewöllen. Michael Imhof Verlag, Fulda. Lambert, O., Pouliquen, H., Larhantec, M., Thorin, C., L'Hostis, M., 2007. Exposure of raptors and waterbirds to anticoagulant rodenticides (difenacoum, bromadiolone, coumatetralyl, coumafen, brodifacoum): epidemiological survey in loire atlantique (France). Bull. Environ. Contam. Toxicol. 79, 91–94. Langenbach, A.A., 1982. Die Beute der Schleiereule. Falke 29, 193–196. López-Perea, J.J., Camarero, P.R., Molina-López, R.A., Parpal, L., Obón, E., Solá, J., Mateo, R., 2015. Interspecific and geographical differences in anticoagulant rodenticide residues of predatory wildlife from the Mediterranean region of Spain. Sci. Total Environ. 511, 259–267. Love, R., Webbon, C., Glue, D.E., Harris, S., 2000. Changes in the food of British barn owls (Tyto alba) between 1974 and 1997. Mammal Rev. 30, 107–129. Lund, M., 1964. Resistance to warfarin in the common rat. Nature 203, 778. Marti, C.D., 1973. Food consumption and pellet formation rates in four owl species. Wilson Bulletin 85, 178–181. Morris, P., 1979. Rats in the diet of the barn owl (Tyto alba). J. Zool. 189, 540–545. Murray, M., 2011. Anticoagulant rodenticide exposure and toxicosis in four species of birds of prey presented to a wildlife clinic in Massachusetts, 2006–2010. Journal of Zoo and Wildlife Medicine 42, 88–97. Newton, I., Wyllie, I., Freestone, P., 1990. Rodenticides in British barn owls. Environ. Pollut. 68, 101–117. Newton, I., Wyllie, I., Gray, A., Eadsforth, C.V., 1994. The toxicity of the rodenticide flocoumafen to barn owls and its elimination via pellets. Pestic. Sci. 41, 187–193. Newton, I., Wyllie, I., Dale, L., 1997. Mortality causes in British Barn Owls (Tyto alba), based on 1,101 carcasses examined during 1963–1996. General Technical Report NC 1997. United States Department of Agriculture Forest Service, Winnipeg, Manitoba, Canada, pp. 299–307. Obuch, J., Benda, P., 2009. Food of the barn owl (Tyto alba) in the Eastern Mediterranean. Slovak Raptor Journal 3, 41–50. Pelz, H.J., 2007. Spread of resistance to anticoagulant rodenticides in Germany. Int. J. Pest Manag. 53, 299–302.

157

RCoreTeam, 2013. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing. Sage, M., Coeurdassier, M., Defaut, R., Gimbert, F., Berny, P., Giraudoux, P., 2008. Kinetics of bromadiolone in rodent populations and implications for predators after field control of the water vole, Arvicola terrestris. Sci. Total Environ. 407, 211–222. Schenke, D., Broll, A., Esther, A., Jacob, J., Schmolz, E., 2013. Analysis of anticoagulant rodenticides in liver using solid supported liquid extraction and LC-ESI/MS/MS. SETAC Europe 23rd Annual Meeting. Society of Environmental Toxicology and Chemistry Europe (SETAC Europe), Glasgow, UK, pp. 157–158 Abstract Book. Shore, R.F., Birks, J.D.S., Afsar, A., Wienburg, C.L., Kitchener, A.C., 2003. Spatial and temporal analysis of second-generation anticoagulant rodenticide residues in polecats (Mustela putorius) from throughout their range in Britain, 1992–1999. Environ. Pollut. 122, 183–193. Smith, D.G., Wilson, C.R., Frost, H.H., 1972. Seasonal food habits of barn owls in Utah. Great Basin Naturalist 32, 229–234. Stone, W.B., Okoniewski, J.C., Stedelin, J.R., 1999. Poisoning of wildlife with anticoagulant rodenticides in New York. J. Wildl. Dis. 35, 187–193. Thomas, P.J., Mineau, P., Shore, R.F., Champoux, L., Martin, P.A., Wilson, L.K., Fitzgerald, G., Elliott, J.E., 2011. Second generation anticoagulant rodenticides in predatory birds: probabilistic characterisation of toxic liver concentrations and implications for predatory bird populations in Canada. Environ. Int. 37, 914–920. Tosh, D.G., McDonald, R.A., Bearhop, S., Lllewellyn, N.R., Fee, S., Sharp, E.A., et al., 2011. Does small mammal prey guild affect the exposure of predators to anticoagulant rodenticides? Environ. Pollut. 159, 3106–3112. Tosh, D.G., McDonald, R.A., Bearhop, S., Llewellyn, N.R., Montgomery, W.I., Shore, R.F., 2012. Rodenticide exposure in wood mouse and house mouse populations on farms and potential secondary risk to predators. Ecotoxicology 21, 1325–1332. Umweltbundesamt, 2014. Nagetierbekämpfung mit Antikoagulanzien - Antworten auf häufig gestellte Fragen. Umweltbundesamt, Berlin, Germany Available from http:// www.umweltbundesamt.de/sites/default/files/medien/378/publikationen/faq_ rodentizide_3.auflage.pdf (accessed June 2015). Walker, L.A., Llewellyn, N.R., Peréira, M.G., Potter, E.D., Molenaar, F.M., Sainsbury, A.W., Shore, R.F., 2010. Anticoagulant Rodenticides in Predatory Birds 2007 & 2008: A Predatory Bird Monitoring Scheme (PBMS) Report (19 pp.). Wood, S., 2007. The mgcv package. Available from https://cran.r-project.org/web/ packages/mgcv/mgcv.pdf (accessed July 2015).