**2.2 Animals monitoring and sampling**

Concerning such rare species, it is particularly difficult to obtain sufficient sample material from enough individuals to support analysis and statistics. For ethical reasons it was not imaginable to trap or kill otters or ospreys for analyses. To avoid any vital risk related to handling, capture and bleed of animals were not considered. Furthermore, otter and osprey are fully protected by national and international laws, and listed as species of interest by the European Community (Habitats Directive 92/43/EC, Birds Directive 79/409/CEE). All operations were therefore entirely conducted under appropriate authorizations by a noninvasive approach. A large network, constituted by people in charge of otter and osprey studying and monitoring in mainland France was built to organize and enhance sampling under the coordination of the Muséum d'Orléans. The national agency for game and wildlife (ONCFS), hunting federations (FDC), the national agency for water and aquatic environments (ONEMA), health centres of the national union (UFCS) and of the birds protection league (LPO – French representative of Bird Life International), the national research centre on birds population biology (CRBPO, associated with the French national museum of natural history MNHN and Mr Rolf Wahl, in charge of osprey ringing program in France), the French Ministry of Environment (MEEDDM and DREAL Centre), the national agency for forests (ONF), private land owners and companies, museums, associations ("Loiret Nature Environnement") and regional naturalists were contributors for this study.

Concerning otters, only road-traffic killed individuals and those found dead in the wild in study area were collected. Based on visual observation, carcasses found more than 24h (during summer) or 48h (during winter) after road collision were considered too degraded

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

Food Webs: The Case of Eurasian Otter (*Lutra lutra*) and Osprey (*Pandion haliaetus*) in Loire … 295

Lindane (*gamma-*HCH) Organochlorine insecticide C6H6Cl6 1998 Endosulfan Organochlorine insecticide C9H6Cl603S 2007 DDT Organochlorine insecticide C14H9Cl5 1972 Heptachlor Organochlorine insecticide C10H5Cl7 1973 Aldrin Organochlorine insecticide C12H8Cl6 1992 Methoxychlor Organochlorine insecticide C16H15Cl3O2 2002

Carbofuran Carbamate insecticide C12H15NO3 2008 Mevinphos Organophosphate insecticide C7H13O6P 2004 Phorate Organophosphate insecticide C7H1702PS3 2004 Dichlorvos Organophosphate insecticide C4H7Cl2O4P 2007 Terbufos Organophosphate insecticide C9H2102PS3 2004 Diazinon Organophosphate insecticide C12H21N2O3PS Still in use Disulfoton sulfone Organophosphate insecticide C8H1902PS3 2004 Chlorpyriphos ethyl Organophosphate insecticide C9H11Cl3NO3PS Still in use Fenitrothion Organophosphate insecticide C9H12NO5PS Still in use Pyrimiphos methyl Organophosphate insecticide C11H20N3O3PS Still in use Malathion Organophosphate insecticide C10H1906PS2 2008 Fenthion Organophosphate insecticide C10H1503PS2 2005 Parathion Organophosphate insecticide C10H14NO5PS 2002 Methidathion Organophosphate insecticide C6H11N2O4PS3 2004 Triazophos Organophosphate insecticide C12H16N303PS 1992 Trifluraline Anilide herbicide C13H16F3N3O4 2008 Atrazine Triazine herbicide C8H14ClN5 2003 Simazine Triazine herbicide C7H12CLN5 2003 Terbuthylazine Triazine herbicide C9H16ClN5 2003 Cyanazine Triazine herbicide C9H13ClN6 2004 Alachlor Chloroacetanilide herbicide C14H20ClNO2 2008 Metolachlor Organochlorine herbicide C15H33ClNO2 2003

Methiocarb Carbamate insecticide,

Diuron Substituted phenylurea

analyzed in this study.

Pesticides Pesticides family and main use Molecular formula Date of ban in France (or

molluscicide C11H15NO2S Still in use

herbicide C9H10Cl2N2O 2008

Epoxyconazole Fongicide C17H13ClFN3O Still in use Tefluthrine Pyrethroid insecticide C17H14ClF7O2 Still in use Cyhalothrine Lambda Pyrethroid insecticide C23H19ClF3NO3 Still in use Permethrine Cis Pyrethroid insecticide C21H20Cl2O3 Still in use Cyfluthrine 2 Pyrethroid insecticide C22H18Cl2FNO3 Still in use Cypermethrine 2 Pyrethroid insecticide C22H19CL2NO3 Still in use Fenvalerate Cis Pyrethroid insecticide C25H22ClNO3 Still in use Deltamethrine Pyrethroid insecticide C22H19Br2NO3 Still in use

Table 1. List of families and uses, molecular formulae, current status of the compounds

current status)

and not taken into account for post-mortem examination and toxicological analyses. Concerning ospreys, non-hatched eggs and dead young in nests were collected during chicks ringing operations. As scientists and birdwatchers monitor a majority of osprey nests in continental France, non-hatched eggs and dead young in nests were reported and sampled as soon as possible. France is also a major crossing area for migrating osprey from different populations (Hake et al. 2001; Dennis, 2008; Strandberg et al. 2009). Due, in one way, to the extreme rarity of this species in continental France (less than one hundred reproductive individuals), and in an other way that "foreigners" individuals (*i.e.* born in neighbour countries, but potentially breeders in France) are able to be found dead within the national territory (naturally or after illegal shots, electrocution on power cables, or drown in fish farms), migrating individuals flying towards reproduction areas elsewhere in Europe (Germany, Great-Britain, Scandinavia) completed sampling.

All samples were deep-frozen (-40 °C) prior to analyses. For each otter or osprey carcass, a necropsy was performed, and about 20 g of liver was sampled. This organ was preferred to fat because some otters and a lot of ospreys have very little fat, particularly at the end of spring migration concerning these latter. Otter sex and weight were determined; animals were measured (total and head, body, foot and tail lengths). Age was defined as "juvenile" (milk teeth, little size and weight), "subadult" (adult size and weight, teeth without wear and tartar) and "adult" or "old" (worn teeth with tartar). Body condition index *K* was determined according to Kruuk and Conroy (1991). Osprey sex and weight were determined; animals were measured (wing, body, foot and tail lengths). Non-hatched osprey eggs were drilled and emptied; eggshell was conserved for future studies on shell thickness. Age was defined as "egg" (non-hatched), "juvenile" (non-flying hatched individual), "subadult" (emancipated individual with the characteristic creamy fringe on feathers) and "adult" (adult size and plumage) (Dennis, 2008). Each animal (otter or osprey) is characterised by a specific case-record gathering discovery circumstances, clinical and biometrical data. After necropsies, carcasses were conserved for further showing or collection in museums or, if too degraded, systematically destroyed according to law.

#### **2.3 Choice of compounds**

Pesticides uses in France are one of the biggest in the world. Various compounds have been used for wood, vineyards, orchard, crops or ornamental plant protection, human or livestock health, and roads, railways or boat maintenance. Origins and flow of compounds are complex and aquatic habitats are exposed to both direct and indirect contamination.

In such a generalist approach, the choice of analyzed compounds is crucial and has to be representative:


A specific detection and quantification methodology was developed for pesticides in the toxicology laboratory of the college of veterinary medicine (VetAgro Sup, Lyon, France) during routine analyses on wild, game or domestic fauna. Compounds were chosen according to their toxicity on fauna, persistence in soil and water and accumulation in food webs. Regular complements and upgrades were added, as a function of new compounds or new detection techniques. Detected compounds are listed in Table 1 below.

#### Semi Aquatic Top-Predators as Sentinels of Diversity and Dynamics of Pesticides in Aquatic Food Webs: The Case of Eurasian Otter (*Lutra lutra*) and Osprey (*Pandion haliaetus*) in Loire … 295

294 Pesticides in the Modern World - Risks and Benefits

and not taken into account for post-mortem examination and toxicological analyses. Concerning ospreys, non-hatched eggs and dead young in nests were collected during chicks ringing operations. As scientists and birdwatchers monitor a majority of osprey nests in continental France, non-hatched eggs and dead young in nests were reported and sampled as soon as possible. France is also a major crossing area for migrating osprey from different populations (Hake et al. 2001; Dennis, 2008; Strandberg et al. 2009). Due, in one way, to the extreme rarity of this species in continental France (less than one hundred reproductive individuals), and in an other way that "foreigners" individuals (*i.e.* born in neighbour countries, but potentially breeders in France) are able to be found dead within the national territory (naturally or after illegal shots, electrocution on power cables, or drown in fish farms), migrating individuals flying towards reproduction areas elsewhere in Europe

All samples were deep-frozen (-40 °C) prior to analyses. For each otter or osprey carcass, a necropsy was performed, and about 20 g of liver was sampled. This organ was preferred to fat because some otters and a lot of ospreys have very little fat, particularly at the end of spring migration concerning these latter. Otter sex and weight were determined; animals were measured (total and head, body, foot and tail lengths). Age was defined as "juvenile" (milk teeth, little size and weight), "subadult" (adult size and weight, teeth without wear and tartar) and "adult" or "old" (worn teeth with tartar). Body condition index *K* was determined according to Kruuk and Conroy (1991). Osprey sex and weight were determined; animals were measured (wing, body, foot and tail lengths). Non-hatched osprey eggs were drilled and emptied; eggshell was conserved for future studies on shell thickness. Age was defined as "egg" (non-hatched), "juvenile" (non-flying hatched individual), "subadult" (emancipated individual with the characteristic creamy fringe on feathers) and "adult" (adult size and plumage) (Dennis, 2008). Each animal (otter or osprey) is characterised by a specific case-record gathering discovery circumstances, clinical and biometrical data. After necropsies, carcasses were conserved for further showing or collection in museums or, if too degraded, systematically destroyed according to law.

Pesticides uses in France are one of the biggest in the world. Various compounds have been used for wood, vineyards, orchard, crops or ornamental plant protection, human or livestock health, and roads, railways or boat maintenance. Origins and flow of compounds are complex and aquatic habitats are exposed to both direct and indirect contamination. In such a generalist approach, the choice of analyzed compounds is crucial and has to be


A specific detection and quantification methodology was developed for pesticides in the toxicology laboratory of the college of veterinary medicine (VetAgro Sup, Lyon, France) during routine analyses on wild, game or domestic fauna. Compounds were chosen according to their toxicity on fauna, persistence in soil and water and accumulation in food webs. Regular complements and upgrades were added, as a function of new compounds or


new detection techniques. Detected compounds are listed in Table 1 below.


(Germany, Great-Britain, Scandinavia) completed sampling.

**2.3 Choice of compounds** 

representative:

species.


Table 1. List of families and uses, molecular formulae, current status of the compounds analyzed in this study.

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

was based on retention times and 3 or 4 ions.

**2.4.4 Herbicides analyses** 

**3. Results** 

**3.1.1 Otters** 

Food Webs: The Case of Eurasian Otter (*Lutra lutra*) and Osprey (*Pandion haliaetus*) in Loire … 297

(0.32 mm ID, 0.25µm film) was used. For each samples standard and spiked sample, 2 µL were injected. The temperature program was common to OCs', PCBs and pyrethroids (initial temp: 100°C, first ramp 6°C/min up to 220 °C held for 10 min, 2nd ramp 7 °C/min up to 285°C, held for 1 min, total run time 42.29 min) Injector was at 230°C, detector at 300°C. Total He flow was 9 ml/min. Pyrethroids were identified according to their retention times. Linearity was confirmed between 10 and 100 ng.g-1 with 5 point calibration curves and r2 >0.99. Recovery was determined between 82% and 94% for all spiked samples and repeatability was considered acceptable with coefficients of variation <15%. For all positive samples, a confirmatory analysis was performed with GC/MS in SIM mode. Identification

2 g of muscle sample was shaken during 5 minutes in 8 ml acetone, and then centrifuged at 4x g; supernatant was placed in separate tubes, and this extraction was performed twice. Samples were evaporated under nitrogen, and dry extract was dissolved in 1 ml aceton/methanol (50:50) solution. Extract was then purified with a SPE C18 500 mg column conditioned with 2 ml acetone and 2 ml methanol. Column was vacuum dried and purified samples were diluted in 3 ml acetone. After drying under nitrogen, samples were diluted in 1 ml methanol. Herbicides (Trifluraline, Atrazine, Simazine, Terbuthylazine, Diuron, Alachlor, Metolachlor, Cyanazine, Epoxyconazloe) concentrations were determined by GC/MS spectrometry. A 5973N MS coupled with a 6890 GC (Agilent®) was used, with a 30m HP5-MS column (0.25 mm ID, 0.25µm thickness). For each samples standard and spiked sample, 2 µL were injected. The temperature program was 85°C held 1 min, followed by 6°C/min up to 170°C (held for 12 min), then followed by 20°C/min up to 280°C, held for 4.33 min (total run time 37 min). Injector was at 250°C and in the splitless mode. Each herbicide was identified based on the following criteria: retention time and 3-4 fragmentation ions with pre-defined relative amounts and 20% variability acceptance for each ion. Linearity was confirmed between 100 and 500 ng.g-1 with 5 point calibration curves and r2 >0.99. Recovery was determined between 67% and 98% for all spiked samples

and repeatability was considered acceptable with coefficients of variation <15%.

two variables. Statistics were performed using *R.* (Ihaka and Gentleman 1996).

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

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

**2.5 Calculation methods and statistical analysis** 

**3.1 General characteristics of sampled material** 

#### **2.4 Pesticides quantification methods 2.4.1 Organochlorine pesticides**

2.0-8.0 g of tissue were sampled and 30 ml of hexane/acetone 75/25 mix was added. Each sample was blended twice with an Ultraturrax® (Ika, Werke, Germany) and filtered trough a phase separator membrane. The extract was evaporated at 60 °C in a rotary evaporator. The dry extract was dissolved in 10 ml hexane.

Two ml of fuming sulphuric acid (SO3 7%) were added, and after centrifugation at 4x *g*, 1 ml of the supernatant was used for OC pesticides quantification by gas chromatography with electron capture detection material. Temperature program and injection conditions are described in Lemarchand et al. (2007; 2010). Each sample was run in duplicate. Organochlorine pesticides concentrations were calculated by using different mix standards. Recovery level on standard mixtures was always greater than 92%. All standards were purchased from CIL (St Foy la Grande, France), and purity was > 99%. Linearity was determined between 5 and 100 ng.g-1 (*r*² > 0.99 on standards and spiked samples, 5-point calibration curves). Limits of detection were between 0.5 and 1.0 ng.g-1 lipids for individual PCB congeners. Cod liver oil (BCR349) certified material was used as a regular quality control.
