**2.5 Calculation methods and statistical analysis**

Geometric means of *p,p'*-DDE, *p,p'-*DDD and *p,p'-*DDT were added to calculate the sum of DDTs (Σ DDTs). Geometric means of lindane, endosulfan, DDE, DDD, DDT, heptachlor, heptachlor epoxyde, aldrin and metoxychlor were summed to provide the sum of pesticide concentrations (Σ Pesticides). All these were chosen by the National Veterinary School of Lyon (VetAgro Sup, France) standard protocol (Mazet et al. 2005; Lemarchand et al. 2007, 2010). The Mann-Whitney test was used to compare two independent samples, Kruskall-Wallis for *k* comparisons, Spearman correlation rank test to quantify associations between two variables. Statistics were performed using *R.* (Ihaka and Gentleman 1996).

## **3. Results**

#### **3.1 General characteristics of sampled material**

#### **3.1.1 Otters**

Otters have been systematically collected since the beginning of the toxicological program along Loire River and tributaries catchment (2004). This program allowed an increase in the

Semi Aquatic Top-Predators as Sentinels of Diversity and Dynamics of Pesticides in Aquatic

Lemarchand, 2007).

**3.1.2 Ospreys** 

Food Webs: The Case of Eurasian Otter (*Lutra lutra*) and Osprey (*Pandion haliaetus*) in Loire … 299

attributed to the larger territory of males compared to females, with associated higher risk of vehicular collision during food or new habitat foraging, particularly in the case of natural recolonization of unknown habitats (Foster-Turley et al. 1990; Rosoux and Tournebize, 1993; Kruuk, 2006; Lemarchand, 2007). Most of dead otters (30 out of 51, *i.e.* 59%) were adults, 11 were subadults, 9 were juvenile and only one was an old individual. This mortality picture is different from those observed in previous studies, where most of discovered otters were juvenile or subadults, with an linear increasing in probability of death with age, considering the rareness of very old individuals in nature (Kruuk and Conroy, 1991; Kruuk, 2006;

With the exception of only two individuals found in the wild during the study, all otters died after a vehicular collision. These results tend to confirm that road casualties seem to be one of the main causes of mortality of otters, but, as suggested by Kruuk and Conroy (1991), and Kruuk (2006), it is the easiest way to find dead otters, those dying of other causes in the wild having far less probability of being found. This bias of carcasses collect was difficult to overlap, because of the huge human and financial costs of systematic search on riverbanks, ponds and lakes in such a study area. Considering this bias of collect, assessing the real hierarchy of causes of mortality remains hard for such a species. Among those found in the wild, one was an adult found dead without any clinical sign or injury, the other was a very little otter, only a few days aged, died of starvation after the dead or the abandon by its mother. This finding was surprising: as young otters live all their time in den, the probabilities of discover them if they die is very low (Kruuk, 2006). With the exception of various injuries caused by road collisions, all otters were in good physical conditions, with no apparent organ damage due to intoxication, like hemorrhages, organ abnormality or wound. Body condition index was systematically comprised between 0,5 and 1,4: it can be assumed that none otter collected in this study was in poor health condition or particularly fat (index < 0,5 or > 1,4, respectively; Kruuk, 2006). Medium body condition index of all otters was 0,99, very close to the value of reference (=1; Kruuk, 2006). Differences between body condition indexes K, total length or weight of otters from upper or lower part of Loire River catchment were not significant. Post-mortem examinations of otters never showed any clinical sign of severe intoxication, like organ or tissue abnormality, secretions, hemorrhages or anemia. Lead pellets were found on two occasions in carcasses but were not a death causal agent. According to Bo Madsen et al. (2000) or Simpson et al. (2005), otters are generally few concerned by natural intoxication (*e.g.* botulism), viral or bacterial diseases. Individuals examined here never showed any strong disease or natural intoxication signs.

Osprey population in mainland France is monitored since the natural come back of the species as breeding one in 1984 (synthesised in Nadal and Tariel, 2008). From only one in 1984, population of breeding ospreys in the study area increased to 35 active nests in 2010, the overall breeding success during 1985-2006 periods was 2.0 fledglings per active nest. This value is higher than the stable population threshold (=0,8), and, associated with the recorded survival rate of adults of 0.97, suggests a very good reproduction dynamics of osprey in the study area during this period (Poole, 1989; Rattner et al. 2004; Wahl and Tariel, 2006; Dennis, 2008; Nadal and Tariel, 2008). The large, favourable and non-fully occupied potential habitats along the Loire River, associated with an important and diversified food resource were the main factors of this reproductive success. Ospreys have been

scientific use of previously collected and stocked individuals, for toxicological analyses first, but also for genetic study of otter recolonization, diet, biometry or causes of mortality approaches (Mucci et al. 2010). Main characteristics of otters analyzed in this study are summarized in table 2.


Table 2. Main characteristics and causes of death of otters in this study.

Fifty-one otters were necropsied and analyzed for this study, with 24 females (47%) and 27 males (53%). Sex-ratio of the sample was very close to the equilibrium but characterized by a slight over-representation of males. This was noted in previous studies, and was generally attributed to the larger territory of males compared to females, with associated higher risk of vehicular collision during food or new habitat foraging, particularly in the case of natural recolonization of unknown habitats (Foster-Turley et al. 1990; Rosoux and Tournebize, 1993; Kruuk, 2006; Lemarchand, 2007). Most of dead otters (30 out of 51, *i.e.* 59%) were adults, 11 were subadults, 9 were juvenile and only one was an old individual. This mortality picture is different from those observed in previous studies, where most of discovered otters were juvenile or subadults, with an linear increasing in probability of death with age, considering the rareness of very old individuals in nature (Kruuk and Conroy, 1991; Kruuk, 2006; Lemarchand, 2007).

With the exception of only two individuals found in the wild during the study, all otters died after a vehicular collision. These results tend to confirm that road casualties seem to be one of the main causes of mortality of otters, but, as suggested by Kruuk and Conroy (1991), and Kruuk (2006), it is the easiest way to find dead otters, those dying of other causes in the wild having far less probability of being found. This bias of carcasses collect was difficult to overlap, because of the huge human and financial costs of systematic search on riverbanks, ponds and lakes in such a study area. Considering this bias of collect, assessing the real hierarchy of causes of mortality remains hard for such a species. Among those found in the wild, one was an adult found dead without any clinical sign or injury, the other was a very little otter, only a few days aged, died of starvation after the dead or the abandon by its mother. This finding was surprising: as young otters live all their time in den, the probabilities of discover them if they die is very low (Kruuk, 2006). With the exception of various injuries caused by road collisions, all otters were in good physical conditions, with no apparent organ damage due to intoxication, like hemorrhages, organ abnormality or wound. Body condition index was systematically comprised between 0,5 and 1,4: it can be assumed that none otter collected in this study was in poor health condition or particularly fat (index < 0,5 or > 1,4, respectively; Kruuk, 2006). Medium body condition index of all otters was 0,99, very close to the value of reference (=1; Kruuk, 2006). Differences between body condition indexes K, total length or weight of otters from upper or lower part of Loire River catchment were not significant. Post-mortem examinations of otters never showed any clinical sign of severe intoxication, like organ or tissue abnormality, secretions, hemorrhages or anemia. Lead pellets were found on two occasions in carcasses but were not a death causal agent. According to Bo Madsen et al. (2000) or Simpson et al. (2005), otters are generally few concerned by natural intoxication (*e.g.* botulism), viral or bacterial diseases. Individuals examined here never showed any strong disease or natural intoxication signs.

#### **3.1.2 Ospreys**

298 Pesticides in the Modern World - Risks and Benefits

scientific use of previously collected and stocked individuals, for toxicological analyses first, but also for genetic study of otter recolonization, diet, biometry or causes of mortality approaches (Mucci et al. 2010). Main characteristics of otters analyzed in this study are

LF 01 Female Juvenile 1,21 Upper part Collision LM 123 Male Subadult 1,07 Lower part Collision LM 02 Male Adult 0,91 Upper part Collision LM 124 Male Subadult 1,00 Lower part Collision LM 05 Male Adult 1,08 Upper part Collision LM 125 Male Adult 1,04 Lower part Collision LM 09 Male Subadult 1,11 Upper part Collision LM 126 Male Adult 1,04 Lower part Collision LM 12 Male Subadult 0,71 Upper part Collision LM 127 Male Subadult 1,10 Lower part Collision LM 13 Male Adult 0,91 Upper part Collision LM 129 Male Adult 1,16 Lower part Collision LM 14 Male Adult 1,07 Upper part Collision LM 130 Male Juvenile 1,17 Lower part Collision LM 16 Male Juvenile 0,62 Upper part Natural LM 141 Male Juvenile 0,85 Lower part Collision YL Male Juvenile 0,82 Upper part Starvation LM 144 Male Adult 0,89 Lower part Collision LF 62 Female Juvenile 0,95 Lower part Collision LM 148 Male Adult 1,20 Lower part Collision LF 64 Female Adult 1,01 Lower part Collision LM 153 Male Adult 1,36 Lower part Collision LF 68 Female Juvenile 0,64 Lower part Collision LM 154 Male Subadult 1,17 Lower part Collision LM 71 Male Adult 1,28 Lower part Collision LM 156 Male Juvenile 1,12 Lower part Collision LF 72 Female Adult 1,06 Lower part Collision LF 157 Female Subadult 0,82 Lower part Collision LF 74 Female Subadult 0,97 Lower part Collision LF 158 Female Subadult 0,58 Lower part Collision LF 77 Female Old 0,67 Lower part Collision LF 159 Female Juvenile 1,03 Lower part Collision LF 78 Female Adult 0,71 Lower part Collision LM 160 Male Adult 1,20 Lower part Collision LM 83 Male Adult 0,68 Lower part Collision LF 161 Female Adult 0,95 Lower part Collision LF 85 Female Subadult 0,96 Lower part Collision LF 162 Female Adult 0,98 Lower part Collision LM 86 Male Adult 1,39 Lower part Collision LF 163 Female Adult 0,83 Lower part Collision LF 88 Female Adult 1,02 Lower part Collision LF 164 Female Adult 0,91 Lower part Collision LF 89 Female Subadult 0,96 Lower part Collision LF 165 Female Adult 0,95 Lower part Collision LM 90 Male Adult 0,89 Lower part Collision LF 167 Female Adult 0,92 Lower part Collision LF 91 Female Adult 1,03 Lower part Collision LF 168 Female Adult 1,06 Lower part Collision LF 93 Female Adult 0,82 Lower part Collision LF 169 Female Adult 1,05 Lower part Collision LM 94 Male Adult 1,38 Lower part Collision

death Otter Sex Age

Body index K

Catchment Origin

Cause of death

Cause of

summarized in table 2.

Body index K

Catchment Origin

Table 2. Main characteristics and causes of death of otters in this study.

Fifty-one otters were necropsied and analyzed for this study, with 24 females (47%) and 27 males (53%). Sex-ratio of the sample was very close to the equilibrium but characterized by a slight over-representation of males. This was noted in previous studies, and was generally

Otter Sex Age

Osprey population in mainland France is monitored since the natural come back of the species as breeding one in 1984 (synthesised in Nadal and Tariel, 2008). From only one in 1984, population of breeding ospreys in the study area increased to 35 active nests in 2010, the overall breeding success during 1985-2006 periods was 2.0 fledglings per active nest. This value is higher than the stable population threshold (=0,8), and, associated with the recorded survival rate of adults of 0.97, suggests a very good reproduction dynamics of osprey in the study area during this period (Poole, 1989; Rattner et al. 2004; Wahl and Tariel, 2006; Dennis, 2008; Nadal and Tariel, 2008). The large, favourable and non-fully occupied potential habitats along the Loire River, associated with an important and diversified food resource were the main factors of this reproductive success. Ospreys have been

Semi Aquatic Top-Predators as Sentinels of Diversity and Dynamics of Pesticides in Aquatic

**3.2 Contamination by organochlorine pesticides** 

immediate threat to otter conservation.

OC pesticides Males Females Juveniles Sub

organochlorine pesticides (mg.kg-1).

Food Webs: The Case of Eurasian Otter (*Lutra lutra*) and Osprey (*Pandion haliaetus*) in Loire … 301

Results concerning contamination of otters by OC pesticides are represented in table 4. OC pesticides and especially DDT metabolites were detected in all (100%) of the analyzed otters, confirming the widespread exposure of otter habitat in France to OC pesticides (Colas et al. 2006; Lemarchand et al. 2007, 2010). Mean concentrations of total OC pesticides in otter liver of the whole Loire River catchment reached 2,2 mg.kg-1 lipid weight, without any statistical variations with the geographical origin of the individuals: the increase in concentrations by going downstream observed in the upper part of the catchment (Lemarchand et al. 2007) was not significant at the whole catchment scale. Differences of various OC pesticides with otter age or sex were not significant. DDT was detected in 17 individuals (33%), confirming quite recent uses of this insecticide, banned in 1973 in France. 15 of these 17 DDTcontaminated otters were coming from the lower part of Loire River catchment. However, DDE was the most abundant of the analyzed DDTs metabolites, confirming the general decrease of otter's exposure to DDT and OC compounds in Europe (Mason, 1998). Lindane constituted the most abundant OC pesticide after DDTs, with quite low concentrations. Aldrin, Dieldrin, Heptachlor and Heptachlor epoxide were very low, often close to the detection limits. Methoxychlor and Endosulfan were never detected in otters. Measured OC pesticides concentrations remained below the available thresholds concerning otter survival (Mason and Macdonald 1993 a,b; 1994). Considering the actual population dynamic in France and elsewhere in Europe, OC compounds are not supposed to constitute an

Otters (n= 51) Ospreys (n= 17)

Old Males Females Eggs Juveniles Sub

adults Adults

Adults &

DDT 0,02 0,01 - 0,01 0,02 - - - - - - DDE 1,12 1,85 0,1 1,45 2,01 10,7 0,55 3,40 - - 1,86 DDD 0,54 0,35 - 0,41 0,60 - - - - - - Lindane 0,11 0,08 0,05 0,15 0,12 - - - - - - Methoxychlor - - - - - - - 0,01 0,3 Aldrin 0,05 0,04 - 0,04 0,11 - - - - - - Heptachlor 0,01 0,01 0,01 0,18 0,14 - - - - - - Hepta. epox. 0,01 0,01 0,01 0,01 0,02 - - - - - - Endosulfan - - - - - - - - - - -

Results concerning contamination of ospreys by OC pesticides are presented in table 4. OC pesticides were detected in all but 5 of the sampled ospreys, and maximum Σ OC pesticides concentrations in liver reached 10,7 mg.kg-1 lipid weight. Only DDTs residues (mainly *p,p'*- DDE) and Methoxychlor were found in samples. DDT by itself was never found in ospreys. These two compounds were never found simultaneously in the same samples. Lindane, Aldrin, Heptachlor, Heptachlor epoxide and Endosulfan were never found in samples. Endosulfan is the only OC pesticide never found in otter or osprey samples. Nevertheless, these compounds were noted in previous studies concerning ospreys, particularly concerning Lindane, Aldrin and Heptachlor epoxide (Ewins *et al.* 1999; Henny *et al.* 2003,

adults

Table 4. Contamination of otters and ospreys from the Loire River catchment by

systematically collected in Mainland France from the end of 2007, when toxicological program on otter was extended to this top-predator. Main characteristics of ospreys analyzed in this study are listed in table 3.

17 osprey samples collected since 2007 in France were used. As some of the birds were ringed, or came from known nests, information about age and origin was established for 12 ospreys (70%). 7 osprey samples (3 non hatched eggs, 3 dead *pulli* in nests and one adult) came from the breeding population along Loire River. The other birds were subadults or adults collected during spring or autumn migration: 4 from Germany, 1 from Norway, and 5 non-ringed birds were from unknown origin. 3 ospreys died after electrocution on power cables, 3 after illegal shots in spite of the full protection of the species by law. Drawing of ospreys in fishponds with inadequate protection nets is another cause of osprey mortality currently emerging: 4 individuals died in the same structure during spring 2009 (March 23rd, 24th, 26th and 28th) in eastern France. To minimize this drawing risk, protection nets were recently modified in several fishponds situated along osprey migration corridors. As observed concerning otters, post-mortem examination never showed any showed any clinical sign of severe intoxication, like organ or tissue abnormality, secretions, hemorrhages or anemia. Three *pulli* from the same nest died of starvation during a long period of bad weather conditions. Three cases of feather pitching syndrome were observed, but these specimen were not collected early enough to support toxicological analyses. The rest of the examined individuals were in good physical conditions (normal size and weigh) and did not show any intoxication or disease sign.


Table 3. Ospreys' characteristics and causes of death analyzed in this study. Osprey numbers corresponds to the chronological sampling order.

#### **3.2 Contamination by organochlorine pesticides**

300 Pesticides in the Modern World - Risks and Benefits

systematically collected in Mainland France from the end of 2007, when toxicological program on otter was extended to this top-predator. Main characteristics of ospreys

17 osprey samples collected since 2007 in France were used. As some of the birds were ringed, or came from known nests, information about age and origin was established for 12 ospreys (70%). 7 osprey samples (3 non hatched eggs, 3 dead *pulli* in nests and one adult) came from the breeding population along Loire River. The other birds were subadults or adults collected during spring or autumn migration: 4 from Germany, 1 from Norway, and 5 non-ringed birds were from unknown origin. 3 ospreys died after electrocution on power cables, 3 after illegal shots in spite of the full protection of the species by law. Drawing of ospreys in fishponds with inadequate protection nets is another cause of osprey mortality currently emerging: 4 individuals died in the same structure during spring 2009 (March 23rd, 24th, 26th and 28th) in eastern France. To minimize this drawing risk, protection nets were recently modified in several fishponds situated along osprey migration corridors. As observed concerning otters, post-mortem examination never showed any showed any clinical sign of severe intoxication, like organ or tissue abnormality, secretions, hemorrhages or anemia. Three *pulli* from the same nest died of starvation during a long period of bad weather conditions. Three cases of feather pitching syndrome were observed, but these specimen were not collected early enough to support toxicological analyses. The rest of the examined individuals were in good physical

conditions (normal size and weigh) and did not show any intoxication or disease sign.

Table 3. Ospreys' characteristics and causes of death analyzed in this study. Osprey

numbers corresponds to the chronological sampling order.

Osprey Sex Age Origin Cause of death bbz 4 female adult Germany electrocution Bbz 3 male adult Loire River electrocution Bbz 7 unknown egg Loire River non hatched egg Bbz 8 unknown egg Loire River non hatched egg Bbz 9-11 male juvenile Loire River pullus dead in nest Bbz 12 unknown juvenile Loire River pullus dead in nest Bbz 13 unknown egg Loire River non hatched egg Bbz 14 female subadult Norway illegal shot Bbz 17 male subadult unknown electrocution Bbz 19 male juvenile Loire River pullus dead in nest Bbz 20 female adult unknown illegal shot Bbz 21 male subadult Germany illegal shot Bbz 23 female adult Germany drawn in fish farm Bbz 24 female subadult unknown drawn in fish farm Bbz 25 male subadult unknown drawn in fish farm Bbz 28 male subadult Germany dead in health center Bbz 31 male adult unknown drawn in fish farm

analyzed in this study are listed in table 3.

Results concerning contamination of otters by OC pesticides are represented in table 4. OC pesticides and especially DDT metabolites were detected in all (100%) of the analyzed otters, confirming the widespread exposure of otter habitat in France to OC pesticides (Colas et al. 2006; Lemarchand et al. 2007, 2010). Mean concentrations of total OC pesticides in otter liver of the whole Loire River catchment reached 2,2 mg.kg-1 lipid weight, without any statistical variations with the geographical origin of the individuals: the increase in concentrations by going downstream observed in the upper part of the catchment (Lemarchand et al. 2007) was not significant at the whole catchment scale. Differences of various OC pesticides with otter age or sex were not significant. DDT was detected in 17 individuals (33%), confirming quite recent uses of this insecticide, banned in 1973 in France. 15 of these 17 DDTcontaminated otters were coming from the lower part of Loire River catchment. However, DDE was the most abundant of the analyzed DDTs metabolites, confirming the general decrease of otter's exposure to DDT and OC compounds in Europe (Mason, 1998). Lindane constituted the most abundant OC pesticide after DDTs, with quite low concentrations. Aldrin, Dieldrin, Heptachlor and Heptachlor epoxide were very low, often close to the detection limits. Methoxychlor and Endosulfan were never detected in otters. Measured OC pesticides concentrations remained below the available thresholds concerning otter survival (Mason and Macdonald 1993 a,b; 1994). Considering the actual population dynamic in France and elsewhere in Europe, OC compounds are not supposed to constitute an immediate threat to otter conservation.


Table 4. Contamination of otters and ospreys from the Loire River catchment by organochlorine pesticides (mg.kg-1).

Results concerning contamination of ospreys by OC pesticides are presented in table 4. OC pesticides were detected in all but 5 of the sampled ospreys, and maximum Σ OC pesticides concentrations in liver reached 10,7 mg.kg-1 lipid weight. Only DDTs residues (mainly *p,p'*- DDE) and Methoxychlor were found in samples. DDT by itself was never found in ospreys. These two compounds were never found simultaneously in the same samples. Lindane, Aldrin, Heptachlor, Heptachlor epoxide and Endosulfan were never found in samples. Endosulfan is the only OC pesticide never found in otter or osprey samples. Nevertheless, these compounds were noted in previous studies concerning ospreys, particularly concerning Lindane, Aldrin and Heptachlor epoxide (Ewins *et al.* 1999; Henny *et al.* 2003,

Semi Aquatic Top-Predators as Sentinels of Diversity and Dynamics of Pesticides in Aquatic

bbz 13

organized according to geographic origin of individuals for a better comparison.

Individuals bbz

Disulfoton

CA pesticides uses.

chains;

19

bbz 3

bbz 7

bbz 8

bbz 9-11

avoided methodological bias in pyrethroids pesticides detection.

moment, but is able to raise in the future with increasing uses.

analyses to other systems, like aquatic systems and associate predators.

These results lead to several hypotheses:

**3.4 Contamination by herbicides** 

bbz 12

Food Webs: The Case of Eurasian Otter (*Lutra lutra*) and Osprey (*Pandion haliaetus*) in Loire … 303

bbz 14

Mevinphos - - - - - - - - - - 0,03 - 0,05 - 0,3 - - Phorate - - - - - - - - - - - - - - - 0,02 - Malathion - - - - - - - - - - - - - - - 0,03 - Parathion - - - - - - - 0,4 - - - - - - - 0,8 - Methidathion - - - - - - - - - - - - 0,02 - - 0,02 -

sulfone - - - - - - - - - - - - 0,3 - 0,3 - 0,04 Triazophos - - - - - - - 0,02 0,02 - - - - 0,03 0,03 - - Table 5. Contamination of ospreys by OP and CA pesticides (mg.kg-1 wet weight). Data are

As described above, ospreys were generally in good physical conditions (adequate mass and total body fat) and did not show any OP pesticides poisoning sign (*e.g.* diarrhea, pulmonary oedema, tightened claws, Berny and Gaillet, 2008) during post-mortem examination. Furthermore, some of them were collected during migration flows, and none bird was found with apparent sign of exhaust potentially brought about by contamination consequences. Measured concentrations remained well below toxic doses of cholinesterase inhibitors (documented as about 10 mg.kg-1 ww) and were not death causal agent of these individuals. Low level of concentrations and of contamination cases frequency should not constitute a threat to the population level, taking into account recent restrictions on OP and

Pyrethroids pesticides residues were never found in any otter or osprey samples (data not shown). The good quality and abundance of samples and the efficiency of the method used



Metabolite of pyrethroids (3-phenoxybenzoic acid 3-PBA) was investigated in osprey eggs of the Washington State, USA (Chu et al. 2007) without being found. Complementary studies are needed to precisely evaluate general contamination of fauna by pyrethroids pesticides. Insect-consumers birds in treated areas (*e.g.* Eurasian skylark *Alauda arvensis*, common quail *Coturnix c.*) and their bird-eating predators (*e.g.* Montagu's harrier *Circus pygargus* or western marsh harrier *C. aeruginosus*) could be used as sentinels for an evaluation of direct transfer of pyrethroids through terrestrial systems first, before generalization of

As observed for OP and CA pesticides, contamination of otters and ospreys by the 8 analyzed herbicides was generally low and few diversified. Only two otters (4%) and 7 ospreys (41%) showed detectable herbicides concentrations. Of the 10 herbicides analyzed, Metholachlor was the only herbicide detected in otters, on two occasions and with low concentrations (0,02 and 0,05 mg.kg-1 ww respectively, data not shown). Two herbicides


bbz 28

bbz 21

bbz 23

bbz 4

bbz 20

bbz 24

bbz 25

bbz 17

bbz 31

2008; Toschik *et al.* 2005). This difference could be related to a different exposure of American ospreys to OC pesticides when compared to European ones, resulting in a higher OC pesticides accumulation pattern in the whole American population. Indeed, American ospreys were exposed to OC pesticides without interruption from the beginning of industrial uses until legal ban. In France, DDT and other OC pesticides were banned before the return of the osprey or at the beginning of population expanding, resulting in a lower and decreasing exposure to contaminants. DDE was detected in 4 individuals (24 %), including 2 eggs from 2 different nests along Loire River and 2 adults, one coming from Loire River. We did not observe any significant variations in OC pesticides concentrations with osprey age, sex or origin. DDE concentrations remained quite low (range 0.0 – 10,7 mg.kg-1 lipid weight). These values were comparable to those noted by Rattner et al. (2004) or Henny et al. (2008), and should not be of concern for osprey direct conservation. DDE concentration in available osprey eggs (n=3) reached 0.0, 4,6 and 5,9 mg.kg-1 lipid weight, respectively. Concerning the latter, the measured values were slightly higher than the 4.2 mg.kg-1 (measured in wet weight) eggshell thinning threshold cited in the literature (Wiemeyer *et al.* 1988; Henny *et al.* 2008), but these eggs did not show any shell breakage and were not damaged. Methoxychlor was detected in 8 individuals (47%), with low values (range 0.0 – 0.93 mg.kg-1 lipid weight, see table 1). General Methoxychlor mean reached 0.01 mg.kg-1 ww, far less than noted by Weber et al. (2003) in Germany, where 100% of the sampled ospreys were contaminated by Methoxychlor. Following these authors, we assume this compound is not a direct threat to ospreys.

#### **3.3 Contamination by organophosphate, carbamate and pyrethroids pesticides**

To a general point of view, contamination of otters and ospreys by the 16 highly toxic cholinesterase inhibitors appeared low and scattered, with only few individuals concerned. Only two otters (4%) and 8 ospreys (47%) were characterized by detectable cholinesterase inhibitors concentrations. Among the OP pesticides analyzed, 7 (Mevinphos, Phorate, Malathion, Parathion, Methidathion, Disulfoton sulfone and Triazophos) were quantified in ospreys and are presented in table 5. Only two OP pesticides (Parathion and Methidathion) were detected in otters, with low concentrations (between 0,02 and 0,03 mg.kg-1 ww, data not shown). No statistical comparison could have been made concerning otter contamination by OP pesticides. Other OP pesticides analyzed in otter and osprey tissues were never been detected. Carbamates pesticides (Methiocarb and Carbofuran) were not quantified in otters and ospreys during this study, however they were recently noted in intoxicated red kites (*Milvus milvus*) in France (Berny and Gaillet, 2008). It can be assumed that diet of otters and ospreys (based on fish) is less exposed to carbamates pesticides accumulation that other diet types of some terrestrial predators like red kite.

OP pesticides were only measured in subadult and adult ospreys, and never found in eggs or *pulli* during this study. None of the individuals coming from the nesting population showed any OP pesticides contamination. OP pesticides variations with osprey age, sex or origin were not significant. Triazophos, Disulfoton sulfone and Mevinphos were the most frequently detected compounds in ospreys (n=4, 3 and 3, respectively). Phorate and Malathion were detected in only one individual, characterized by the highest diversity of compounds (4 compounds with also Parathion and Methidathion) and by the highest concentrations of total OP pesticides (0,9 mg.kg-1 ww, see table 5).


Table 5. Contamination of ospreys by OP and CA pesticides (mg.kg-1 wet weight). Data are organized according to geographic origin of individuals for a better comparison.

As described above, ospreys were generally in good physical conditions (adequate mass and total body fat) and did not show any OP pesticides poisoning sign (*e.g.* diarrhea, pulmonary oedema, tightened claws, Berny and Gaillet, 2008) during post-mortem examination. Furthermore, some of them were collected during migration flows, and none bird was found with apparent sign of exhaust potentially brought about by contamination consequences. Measured concentrations remained well below toxic doses of cholinesterase inhibitors (documented as about 10 mg.kg-1 ww) and were not death causal agent of these individuals. Low level of concentrations and of contamination cases frequency should not constitute a threat to the population level, taking into account recent restrictions on OP and CA pesticides uses.

Pyrethroids pesticides residues were never found in any otter or osprey samples (data not shown). The good quality and abundance of samples and the efficiency of the method used avoided methodological bias in pyrethroids pesticides detection.

These results lead to several hypotheses:

302 Pesticides in the Modern World - Risks and Benefits

2008; Toschik *et al.* 2005). This difference could be related to a different exposure of American ospreys to OC pesticides when compared to European ones, resulting in a higher OC pesticides accumulation pattern in the whole American population. Indeed, American ospreys were exposed to OC pesticides without interruption from the beginning of industrial uses until legal ban. In France, DDT and other OC pesticides were banned before the return of the osprey or at the beginning of population expanding, resulting in a lower and decreasing exposure to contaminants. DDE was detected in 4 individuals (24 %), including 2 eggs from 2 different nests along Loire River and 2 adults, one coming from Loire River. We did not observe any significant variations in OC pesticides concentrations with osprey age, sex or origin. DDE concentrations remained quite low (range 0.0 – 10,7 mg.kg-1 lipid weight). These values were comparable to those noted by Rattner et al. (2004) or Henny et al. (2008), and should not be of concern for osprey direct conservation. DDE concentration in available osprey eggs (n=3) reached 0.0, 4,6 and 5,9 mg.kg-1 lipid weight, respectively. Concerning the latter, the measured values were slightly higher than the 4.2 mg.kg-1 (measured in wet weight) eggshell thinning threshold cited in the literature (Wiemeyer *et al.* 1988; Henny *et al.* 2008), but these eggs did not show any shell breakage and were not damaged. Methoxychlor was detected in 8 individuals (47%), with low values (range 0.0 – 0.93 mg.kg-1 lipid weight, see table 1). General Methoxychlor mean reached 0.01 mg.kg-1 ww, far less than noted by Weber et al. (2003) in Germany, where 100% of the sampled ospreys were contaminated by Methoxychlor. Following these authors, we assume

**3.3 Contamination by organophosphate, carbamate and pyrethroids pesticides**  To a general point of view, contamination of otters and ospreys by the 16 highly toxic cholinesterase inhibitors appeared low and scattered, with only few individuals concerned. Only two otters (4%) and 8 ospreys (47%) were characterized by detectable cholinesterase inhibitors concentrations. Among the OP pesticides analyzed, 7 (Mevinphos, Phorate, Malathion, Parathion, Methidathion, Disulfoton sulfone and Triazophos) were quantified in ospreys and are presented in table 5. Only two OP pesticides (Parathion and Methidathion) were detected in otters, with low concentrations (between 0,02 and 0,03 mg.kg-1 ww, data not shown). No statistical comparison could have been made concerning otter contamination by OP pesticides. Other OP pesticides analyzed in otter and osprey tissues were never been detected. Carbamates pesticides (Methiocarb and Carbofuran) were not quantified in otters and ospreys during this study, however they were recently noted in intoxicated red kites (*Milvus milvus*) in France (Berny and Gaillet, 2008). It can be assumed that diet of otters and ospreys (based on fish) is less exposed to carbamates pesticides accumulation that other diet types of some terrestrial

OP pesticides were only measured in subadult and adult ospreys, and never found in eggs or *pulli* during this study. None of the individuals coming from the nesting population showed any OP pesticides contamination. OP pesticides variations with osprey age, sex or origin were not significant. Triazophos, Disulfoton sulfone and Mevinphos were the most frequently detected compounds in ospreys (n=4, 3 and 3, respectively). Phorate and Malathion were detected in only one individual, characterized by the highest diversity of compounds (4 compounds with also Parathion and Methidathion) and by the highest

concentrations of total OP pesticides (0,9 mg.kg-1 ww, see table 5).

this compound is not a direct threat to ospreys.

predators like red kite.


Metabolite of pyrethroids (3-phenoxybenzoic acid 3-PBA) was investigated in osprey eggs of the Washington State, USA (Chu et al. 2007) without being found. Complementary studies are needed to precisely evaluate general contamination of fauna by pyrethroids pesticides. Insect-consumers birds in treated areas (*e.g.* Eurasian skylark *Alauda arvensis*, common quail *Coturnix c.*) and their bird-eating predators (*e.g.* Montagu's harrier *Circus pygargus* or western marsh harrier *C. aeruginosus*) could be used as sentinels for an evaluation of direct transfer of pyrethroids through terrestrial systems first, before generalization of analyses to other systems, like aquatic systems and associate predators.

#### **3.4 Contamination by herbicides**

As observed for OP and CA pesticides, contamination of otters and ospreys by the 8 analyzed herbicides was generally low and few diversified. Only two otters (4%) and 7 ospreys (41%) showed detectable herbicides concentrations. Of the 10 herbicides analyzed, Metholachlor was the only herbicide detected in otters, on two occasions and with low concentrations (0,02 and 0,05 mg.kg-1 ww respectively, data not shown). Two herbicides

Semi Aquatic Top-Predators as Sentinels of Diversity and Dynamics of Pesticides in Aquatic

CNRS.

**6. References** 

93–100.

La Rochelle.

*Environ. Res.* 90: 142-151.

*Toxicological Conference*. Isle of Skye, pp 47-56.

sauvegarde. Catiche productions - Libris, 32p.

*lutra*). *Chemosphere* 78: 785-792.

Bouchardy, C. (1986). La Loutre. Sang de la Terre, 186p.

Food Webs: The Case of Eurasian Otter (*Lutra lutra*) and Osprey (*Pandion haliaetus*) in Loire … 305

This study was financially supported by: European Commission (FEDER), « Plan Loire Grandeur Nature 2007-2013 », Etablissement Public Loire, Agence de l'Eau Loire-Bretagne, French Ministry of Environment (MEEDDM, DREAL Centre, Plan National d'Actions pour le balbuzard pêcheur en France), VetAgro Sup, Ville d'Orléans, Office National des Forêts, Parc naturel régional des Volcans d'Auvergne, Parc interrégional du Marais poitevin, and

Berny, P., Lachaux, O., Buronfosse, T., Mazallon, M. & Gillet, C. (2002). Zebra mussels

Berny, P. (2007). Pesticides and the intoxication of wild animals. *J. vet. Pharmacol. Therap.* 30:

Berny, P. and Gaillet, J.-R. (2008). Acute poisoning of red kites (*Milvus milvus*) in France : data from the SAGIR network. Journal of Wildlife Disease 44 : 417-426. Bo Madsen, A., Dietz, H.H., Henriksen, P. & Clausen, B. (2000). Survey of Danish free-living

Boscher, A., Gobert, S., Guignard, C., Ziebel, J., L'Hoste, L., Gutleb, A.C., Cauchie, H.-M.,

Bouchardy, C., Rosoux, R. & Boulade, Y. (2001). La Loutre d'Europe, histoire d'une

Britton, J.R., Shepherd, J.S., Toms, S. & Simpson, V. (2005). Presence of Carp, *Cyprinus carpio*,

Chanin, P. (2003*a*). Ecology of the European Otter (*Lutra lutra*). Conserving Natura 2000

Chu, S., Henny, C.J., Kaiser, J.L., Drouillard, K.G., Haffner, G.D. & Letcher, R.J. (2007).

Clavero, M., Prenda, J. & Delibes, M. (2003). Trophic diversity of the Otter (*Lutra lutra* L.) in temperate and Mediterranean freshwater habitats. *J. Biogeog.* 30: 761-769. Colas, C, Caurant, F., Rosoux, R. & De Bellefroid, M.D.N. (2006). Recherche de

Dacthal and chlorophenoxy herbicides and chlorothalonil fungicide in eggs of osprey (*Pandion haliaetus*) from the Duwamish-Lake Washington-Puget Sound area

contaminants organiques et métalliques chez la loutre d'Europe (*Lutra lutra*) dans l'ouest de la France. Rapport de synthèse, plan de restauration national. Université de La Rochelle – UMR 6217 CNRS, Université de Bordeaux I – UMR 5472 CNRS, MEDD, DIREN Poitou-Charentes, Région Poitou-Charentes, Ville de

in the diet of the Otter, *Lutra lutra*. *Fish. Manag. Ecol.* 12: 221-223.

Rivers Ecology Series N°10, English Nature, Peterborough.

of Washington state, USA. *Environ Pollut.* 145: 374-381.

(*Dreissena polymorpha*) as indicators of freshwater contamination with lindane.

otters (*Lutra lutra*). A consecutive collection and necropsy of dead bodies. *In*: Conroy, J.W.H., Yoxon, P. & Gutleb, A.C. (ed.) *Proceedings of the First Otter* 

Hoffmann, L. & Schmidt, G. (2010). Chemical contaminants in fish species from rivers in the North of Luxembourg: Potential impact on the Eurasian otter (*Lutra* 

(Terbuthylazine and Alachlor for 5 individuals each) and fungicide Epoxyconazole (in only one case) were quantified in ospreys (see table 6). None of the ospreys from the nesting population in France showed any herbicide or fungicide contamination. Herbicides variations with osprey age, sex or origin were not significant. Herbicides were not found in osprey eggs during this study. It can be underlined a unique case of contamination by fungicide Epoxyconazole (5,64 mg.kg-1 ww, see table 6), detected in an osprey from Germany. This individual did not show any particular intoxication sign. As for OP and CA pesticides, concentrations of herbicides measured in tissues and low frequency of herbicide detection leads to a probable weak impact of these compounds on species' conservation. Nevertheless, herbicides were very rarely searched in ospreys and very few data are available in literature for comparison. Chu et al. (2007) reported contamination of osprey eggs by a Dacthal structural isomer, indicating that some herbicides could be accumulated in ospreys with a potential reproductive impact on populations.


Table 6. Contamination of ospreys by herbicides and fungicides (mg.kg-1 wet weight).
