Fate in organisms and food webs
Two concepts are essential to the understanding of the incorporation and fate of environmental contaminants in living organisms: exposure and bioavailability. Exposure to a contaminant happens when an organism comes into contact with the contaminant. There are different routes of exposure, depending on the organism and its habitat. In vertebrates, exposure occurs through food and water ingestion, inhalation and dermal contact.
Bioavailability of a contaminant is the fraction that can be assimilated following the exposure and consequently transferred, stored and/or metabolised by the organism (Ramade 2007). The bioavailability of a contaminant is highly dependent on its chemical form, and is influenced by environmental factors (chemical composition of the medium, association with organic matter, etc.) and intrinsic traits of the organism (nutritional status, mobility, etc.). In vertebrates, contaminants are incorporated through the digestive epithelia, the respiratory surfaces and through the skin (Walker et al. 2012). Importantly, the fraction of the contaminant that transits through the gastrointestinal tract without assimilation is not considered bioavailable. In vertebrates, absorbed contaminants may travel through the bloodstream and be distributed to different tissues where diverse biochemical processes take place. For instance, enzymatic metabolism or biotransformation, which occurs mainly in the liver, is an important factor determining the fate of OCPs and PCBs via biochemical processes that render them less (detoxification) or more (bioactivation) toxic and may result in the excretion of the formed metabolite(s) (Letcher et al. 2010). On the other hand, trace elements are not biodegradable (except the organic forms to a certain extent), and detoxification in organisms consists of hiding them within proteins such as metallothioneins, or depositing them in insoluble forms in intracellular granules, that are excreted or stored over the short to long term (Walker et al. 2012). Storage also happens for POPs: given their lipophilicity, reserves of organic compounds can be built in lipid tissues, and consequently released during lipid mobilisation (Borgå et al. 2004 ). Bioaccumulation of contaminants occurs when incorporation overcomes excretion mechanisms, resulting in increasing contaminant concentrations in the organism. Bioaccumulation can happen at all levels of the food web, from primary producers to consumers and decomposers. Furthermore, biomagnification occurs when the dietary transfer of a contaminant results in higher concentrations in the consumer than in the food source (Gray 2002). As a consequence, cumulative quantities of contaminants are transferred at each step of the food web, with apex predators exhibiting the highest concentrations (Fig. 3, Ramade 2007). This phenomenon is often exacerbated in aquatic environments, as a result of two important ecological characteristics (Ramade 2007). First, aquatic food webs are generally longer and/or more complex than terrestrial ones, which increases the number of contaminant trophic transfers. Second, organisms at the base of aquatic food webs, i.e. phyto- and zooplankton, have a high ability to concentrate contaminants from the water. The result is that contaminants are highly mobile and easily transferred to upper levels of aquatic food webs. This is particularly worrying for oceanic ecosystems, because, as seen above, the ocean is usually the last sink for several environmental contaminants.
Southern Ocean seabirds: cycle and feeding habits
Due to its abundant stocks of potential prey, the Southern Ocean hosts a tremendous biomass and diversity of wildlife, including huge populations of avian predators. More than 300 million seabirds breed within this ocean (Van Franeker et al. 1997), of which several millions in the TAAF (e.g., Weimerskirch et al. 1989), including rare and endangered species, such as the Amsterdam albatross Diomedea amsterdamensis and the Indian yellow-nosed albatross Thalassarche carteri (IUCN Red List 2011). Different orders of birds have their breeding grounds in the TAAF: the Sphenisciformes (penguins), the Procellariiformes (albatrosses and petrels), the Charadriiformes (skuas, gulls and terns) and the Pelecaniformes (cormorants), with also one representative of the Anseriformes (ducks). The Sphenisciformes and the Procellariiformes dominate the avian community in terms of biomass and species diversity, respectively (Warham 1990, Williams 1995), and include iconic species such as the emperor penguin Aptenodytes forsteri and the wandering albatross Diomedea exulans.
The annual life cycle of these seabirds is characterised by three energetically-demanding processes: reproduction, migration and moult. Usually, these processes are temporally separated, in order to optimise energy expenditure (King 1974).
Reproduction occurs during the productive austral summer for most species, while only a few are winter breeders (e.g., emperor penguins and grey petrels Procellaria cinerea, Table 3), or have long breeding periods of approximately one year (e.g., king penguins Aptenodytes patagonicus and wandering albatrosses). TAAF seabirds have a low reproductive outcome, raising generally one chick per year (Warham 1990, Williams 1995). Migration strategies are highly variable, with some subantarctic species being resident on their breeding islands (e.g., gentoo penguins Pygoscelis papua, Kerguelen cormorants Phalacrocorax verrucosus and terrestrial species, Table 3), while others migrate over different distances (from Southern Ocean to subtropical, tropical and trans-equatorial destinations, up to the Northern Atlantic and Pacific Oceans) between breeding events (e.g., Marchant and Higgins 1990). Moult is also a critical phase of seabirds’ life cycle. Regular feather replacement is essential to the maintenance of thermic insulation and flight performance (e.g., Bridge 2006). Pattern, duration and timing of moult differ widely amongst TAAF seabird species. Notably, flying seabirds replace their feathers sequentially over a protracted period (two-three months to one year, in general, Bridge 2006) during the inter-breeding season. On the other hand, penguins, which are flightless, diving seabirds, have a unique moulting pattern among birds: they renew all their feathers simultaneously, just before or just after the breeding period (Williams 1995). Importantly, penguins fast during moult, since transient reduction in thermal insulation prevents them from going at sea (Groscolas and Cherel 1992, Cherel et al. 1994a).
Since prey availability is spatially and temporally variable in Southern Ocean waters, seabirds have evolved a wide range of contrasting feeding strategies, in order to exploit their heterogeneous environment while limiting between-species competition. Penguins and diving petrels have evolved exceptional diving capacities, exploiting the water column from pelagic to benthic environments, depending on species (Table 3). On the other hand, flying seabirds explore the marine environment horizontally, feeding from neritic (coastal and peri-insular) to oceanic waters (Table 3). Some species of oceanic seabirds can travel thousands of kilometres over short periods (days), encompassing large latitudinal gradients (e.g., Weimerskirch et al. 2014).
How are environmental contaminants deposited into feathers?
Feathers are unique epidermal structures made of beta-keratin (e.g., Stettenheim 2000). Distal areas emerge first from the feather follicle, whereas proximal aspects form sequentially until the structure is complete (Burger and Gochfeld 1997). During growth (two-three weeks, Burger 1993), each feather shaft contains an axial artery connecting the developing cells to the circulatory system. Different bloodborne materials are thus sequestered in the growing feather: structural compounds, such as amino acids (functional deposition) and non-structural chemicals (incidental deposition), such as hormones and environmental contaminants (Burger 2003, Bortolotti 2010, García-Fernández et al. 2013). The blood supply completely atrophies once the feathers are fully grown, leaving them as inert, long-term archives of the sequestered compounds (Burger and Gochfeld 1997). Some trace elements accumulate particularly in feathers, because of their affinity to –SH groups of keratin. This is the case of Hg, with 50% to 90% of the total body burden being excreted in feathers during moult (e.g., Honda et al. 1986, Braune and Gaskin 1987).
Feather Hg concentrations encompass both dietary intake during feather synthesis and accumulated burdens in internal tissues since the last moult (e.g., Furness et al. 1986). In my doctoral work, I have extensively used feathers for Hg measurement in TAAF seabirds, assuming that the plumage is a good proxy of Hg burdens accumulated over the intermoult period. Conversely, the temporal integration of other trace elements and POPs into feathers are poorly known (Agusa et al. 2005, Jaspers et al. 2007).
Why using the stable isotope method?
Since environmental contaminants are incorporated almost exclusively from the diet, assessing seabird feeding strategies is critical to understand their contaminant exposure (e.g., Becker et al. 2002, Anderson et al. 2010, Leat et al. 2013). In order to evaluate the trophic ecology of a large number of species, conventional studies such as satellite tracking and stomach content analysis can be difficult and expensive to realise. Furthermore, they suffer an important bias: they do not directly inform on assimilated food. A powerful alternative to conventional techniques is the stable isotope method. The isotopic technique is based on the concept that an animal’s chemical composition reflects that of its diet in a predictable manner (Kelly 2000). For example, consumers are enriched in 15N relative to their food and consequently stable nitrogen signatures (15N/14N, δ15N) serve as indicators of their trophic position (Vanderklift and Ponsard 2003). By contrast, stable carbon signatures (13C/12C, δ13C) vary little within food webs and, in the marine environment, δ13C values are mainly used to indicate the foraging habitats of predators, including seabirds (Cherel and Hobson 2007, Cherel et al. 2013). The isotopic method has already been validated in the southern Indian Ocean, with δ15N values of seabirds increasing with trophic level (Cherel et al. 2010).
Moreover, Southern Ocean oceanic waters are marked by a well-defined latitudinal δ13C gradient of particulate organic matter (Trull and Armand 2001) that is reflected in the tissue of consumers (Cherel and Hobson 2007, Jaeger et al. 2010a, Quillfeldt et al. 2010). By measuring the δ13C signature of seabirds, it is thus possible to estimate their main foraging zones, depending on the targeted tissues (Fig. 6): based on blood (and feather) δ13C isoscapes,values less than -22.9 ‰ (-21.2 ‰), -22.9 to -20.1 ‰ (-21.2 to -18.3 ‰), and greater than – 20.1 ‰ (-18.3 ‰) are considered to correspond to th e Antarctic, Subantarctic and Subtropical Zones, respectively (Jaeger et al. 2010a). Furthermore, in the vicinity of islands, δ13C signatures can also depict neritic (high δ13C values) vs. oceanic (low δ13C values) consumers (Cherel and Hobson 2007). Importantly, the stable isotope method enables estimating seabirds feeding strategies over different temporal scales, depending on the sampled tissue and its protein turnover rate (Kelly 2000).
The problem of synchronous vs. asynchronous moults
This doctoral work makes use extensively of body feathers to measure Hg in seabirds. The first step of my doctoral work has thus consisted in a methodological investigation on the reliability of body feathers as biomonitoring tissues.
State of the art and hypothesis. Ideally, a monitoring tool must show little within-and between-individual variations in the concentrations of targeted compounds in order to facilitate the statistical description of spatio-temporal trends within a population. Body feathers are generally considered as the best feather type to sample, for ethical reasons, since their removal does not impair flying ability, but also for scientific reasons, since they are more homogeneous and more representatives of the entire plumage than flight or tail feathers (Furness et al. 1986, Burger and Gochfeld 1992, Jaeger et al. 2009). Thus, I have decided to look at the within-individual heterogeneity of stable isotope values and Hg concentrations in body feathers, which has rarely been addressed before (Thompson et al. 1993, Bond and Diamond 2008, Jaeger et al. 2009, 2010b, Brasso et al. 2013). Since environmental contaminants and stable isotopes are deposited in the plumage during feather growth, the timing, duration and pattern of moult are critical in driving contaminant concentrations and isotopic signatures in feathers (e.g., Furness et al. 1986, Thompson et al. 1998a, Kelly 2000). Indeed, changes in foraging habitat or diet during moult lead to large variations in feather δ13C and δ15N values, respectively (Jaeger et al. 2010b), and feathers that grow at different times present different Hg concentrations, as the Hg body pool is progressively depleted during the moult (Furness et al. 1986, Braune and Gaskin 1987). As illustrated in the introduction (section 1.2.2.), most TAAF seabird species moult sequentially over protracted periods, whereas penguins has a synchronous moult. A simultaneous moult occurs also in chicks of all species, which grow new body feathers towards the end of the rearing period. A synchronous growth theoretically means that all body feathers should have the same chemical composition and thus should show identical stable isotope values and contaminant concentrations, i.e. low within-individual variability. In order to test this hypothesis, I compared within-individual differences (four feathers per individual) in stable isotope Hg values in three model species having synchronous (king penguin adults and white-chinned petrel Procellaria aequinoctialis chicks) and asynchronous (Antarctic prion Pachyptila desolata adults) moulting patterns of body feathers (Paper 1 in the Appendix).
Intrinsic vs. extrinsic factors driving variation in seabird contaminant concentrations
Many different factors may drive variation in contaminant concentrations in seabirds (e.g., Burger 1993, Furness 1993, Borgå et al. 2004 ), including:
– intrinsic factors, i.e. those related to physiology, such as individual traits (sex, age, size), body condition (lipid stores), and mechanisms related to the toxicokinetics and toxicodynamics of environmental contaminants (absorption, biotransformation, excretion).
– extrinsic factors, i.e. those related to the environment and the food source, such as trophic position, foraging habitat and external input of environmental contaminants, including atmospheric deposition on the surface of feathers (e.g., Dauwe et al. 2003, Jaspers et al. 2004).
Understanding the sources of variation of environmental contaminant concentrations in seabirds is crucial for using them as bioindicators of environmental contamination. Namely, if intrinsic factors are not controlled, it is impossible to link reliably concentrations found in seabird tissues with environmental processes. Similarly, if extrinsic factors are not understood, spatio-temporal trends could be associated to the wrong geographic or temporal units. During my doctoral work, different factors of variation have been considered, depending on environmental contaminants and species (Table 4). The influence of body condition could not be investigated, because biometric measures were not taken for a large part of the sampled individuals. The influence of external contamination on feathers was not studied, with feathers being vigorously washed in organic solvents, in order to better evaluate intrinsic deposition of environmental contaminants.
Table of contents :
Chapter 1.. Introduction
1.1. Environmental contaminants: origin and fate in marine ecosystems
1.1.1. Definitions and toxic effects in vertebrates
1.1.2. Sources, large-scale movements and role of the World Ocean
1.1.3. Fate in organisms and food webs
1.1.4. Birds as bioindicators: a critical role
1.2. The Southern Ocean and its avian predators
1.2.1. Oceanographic features
1.2.2. Southern Ocean seabirds: cycle and feeding habits
1.3. Main objective and outline of the thesis
Chapter 2.. Methodological approach: sampling blood and feathers and the stable isotope method
2.1. Advantages of the non-destructive sampling of blood and feathers
2.2. How do environmental contaminants partition between internal organs and blood?.
2.3. How are environmental contaminants deposited into feathers?
2.4. Why using the stable isotope method?
2.5. The problem of synchronous vs. asynchronous moults
2.6. The problem of adults vs. chicks
Chapter 3.. What are the main explanatory factors of contaminant exposure and bioaccumulation?
3.1. Intrinsic vs. extrinsic factors driving variation in seabird contaminant concentrations
3.2. Why using the wandering albatross to assess the influence of intrinsic and extrinsic factors on variation in environmental contaminant concentrations?
3.3. Does age affect environmental contaminant concentrations in blood and feathers?
3.3.1. Age-class differences in Hg concentrations in feathers
3.3.2. Adult age: case study of the wandering albatross
3.4. Do other potential physiological differences affect contaminant concentrations in blood and feathers?
3.4.2. Reproductive status
3.5. Influence of extrinsic factors: does feeding ecology explain contaminant exposure?
3.5.1. Between-species variation in Hg exposure: case study of the Kerguelen community
3.5.2. Between-individual variation in contaminant exposure: case study of gentoo penguins and wandering albatrosses
Chapter 4.. Spatial and temporal variation in contaminant transfer to seabirds in the southern Indian Ocean
4.1. Spatial trends of environmental contaminant transfer to seabirds
4.1.1. Spatial variation in Hg transfer to seabirds
4.1.2. Spatial variation in POPs transfer to seabirds.
4.2. Temporal trends of Hg transfer to seabirds: insights from penguin feathers
Chapter 5… Conclusions, Critical evaluation and Perspectives
5.1.1. Highlights of the doctoral work
5.1.2. The main contaminants in the southern Indian Ocean and elsewhere
5.1.3. Are the measured concentrations of concern?
5.1.4. What are the best bioindicator species for biomonitoring?
5.2. Critical evaluation and perspectives
5.2.1. What do blood and feather contaminant concentrations really mean? .
5.2.2. Can we better identify the source of contaminant exposure in seabirds? .
5.2.3. Are contaminants threatening the immune system of TAAF seabirds? .
5.2.4. Monitoring future trends of contamination in the southern Indian Ocean.