research priorities for seabirds: improving conservation and … · 2. at-sea spatial ecology 2.1....

ENDANGERED SPECIES RESEARCHEndang Species Res

Vol. 17: 93–121, 2012doi: 10.3354/esr00419

Published online May 8

INTRODUCTION

Seabirds have been the focus of human observationfor millennia and of scientific study and observationfor centuries (Charles Darwin, Tierra del Fuego, Jan13, 1833: ‘We were heavily labouring, it was cu riousto see how the Albatross with its widely expand -ed wings, glided right up the wind’. http:// darwin-

online.org.uk/). In more recent years, there has beenstrong evidence of steady, and for some species dra-matic, changes in seabird populations in many areas(Schreiber & Burger 2002, Butchart et al. 2004). Whilefor some populations there has been rapid populationgrowth (e.g. northern fulmars Fulmarus glacialis inthe North Sea in the 1980s), many other seabird spe-cies have demonstrated marked declines. Reported

© Inter-Research 2012 · www.int-res.com

*Email: [email protected]

Addresses for all authors are given in Supplement 1 atwww.int-res.com/articles/suppl/n017p093_supp.pdf

REVIEW

Research priorities for seabirds: improvingconservation and management in the 21st century

R. Lewison1,*, D. Oro2, B. J. Godley3, L. Underhill4, S. Bearhop3, R. P. Wilson5, D. Ainley, J. M. Arcos, P. D. Boersma, P. G. Borboroglu, T. Boulinier, M. Frederiksen,

M. Genovart, J. González-Solís, J. A. Green, D. Grémillet, K. C. Hamer, G. M. Hilton,K. D. Hyrenbach, A. Martínez-Abraín, W. A. Montevecchi, R. A. Phillips, P. G. Ryan,

P. Sagar, W. J. Sydeman, S. Wanless, Y. Watanuki, H. Weimerskirch, P. Yorio1Institute of Ecological Monitoring and Management, San Diego State University, San Diego, California 92182-4614, USA

2CSIC UIB, Institut Mediterrani Estudis Avancats IMEDEA, Esporles 07190, Mallorca, Spain3Centre for Ecology & Conservation, University of Exeter, Cornwall Campus, Penryn, Cornwall TR10 9EZ, UK4Department of Zoology, Animal Demography Unit, University of Cape Town, 7701 Rondebosch, South Africa

5Institute of Environmental Sustainability, Swansea University, Institute of Environmental Sustainability,Singleton Park, Swansea SA2 8PP, UK

ABSTRACT: Seabirds are facing a growing number of threats in both terrestrial and marine habi-tats, and many populations have experienced dramatic changes over past decades. Years ofseabird research have improved our understanding of seabird populations and provided a broaderunderstanding of marine ecological processes. In an effort to encourage future research and guideseabird conservation science, seabird researchers from 9 nations identified the 20 highest priorityresearch questions and organized these into 6 general categories: (1) population dynamics, (2)spatial ecology, (3) tropho-dynamics, (4) fisheries interactions, (5) response to global change, and(6) management of anthropogenic impacts (focusing on invasive species, contaminants and pro-tected areas). For each category, we provide an assessment of the current approaches, challengesand future directions. While this is not an exhaustive list of all research needed to address the myr-iad conservation challenges seabirds face, the results of this effort represent an important synthe-sis of current expert opinion across sub-disciplines within seabird ecology. As this synthesis high-lights, research, in conjunction with direct management, education, and community engagement,can play an important role in facilitating the conservation and management of seabird populationsand of the ocean ecosystems on which they and we depend.

KEY WORDS: Seabird conservation · Research priorities · Population dynamics · Threats · Fish-eries interactions · Tropho-dynamics · Climate change · Spatial ecology

Resale or republication not permitted without written consent of the publisher

OPENPEN ACCESSCCESS

http://www.int-res.com/articles/suppl/n017p093_supp.pdf

Endang Species Res 17: 93–121, 2012

seabird declines follow global patterns of biodiversityloss, with between 10 and 50% of species in well-studied higher taxonomic groups currently threatenedwith extinction (MEA 2005, Schipper et al. 2008).Seabirds are long-lived organisms that exhibit highadult survival and a progressive recruitment to thebreeding population as immature individuals nearsexual maturity (Schreiber & Burger 2002). Theybreed mostly in social structures (i.e. colonies, see Jo-vani et al. 2008) commonly in coastal areas, and theytend to be philopatric, especially once recruited to aspecific colony (Milot et al. 2008).

Many factors have been implicated in documentedchanges in seabird populations, e.g. direct harvest,incidental mortality in fisheries, fisheries closuresand changes in fishing practice, pollution, non-nativeand invasive alien species, and changes in preyavailability as a consequence of fisheries or changingocean conditions. It is estimated that one quarter ofmarine fish stocks on which seabirds depend areoverexploited by fisheries (Hilborn et al. 2003,Roberts et al. 2007). Although fisheries serve as apervasive and global driver of change in marineecosystems, the effects of fisheries on seabirds arenot clearly understood (Watermeyer et al. 2008a,b).Coincident with these factors, seabirds also exhibitconsiderable variability in behavior, ecology, life his-tory and demography (Garthe et al 1996, Hamer et al.2002, Weimerskirch 2002) and respond to short-termand long-term changes in ocean conditions (Aebis-cher et al. 1990, Sydeman et al. 2009). Thus, one ofthe central challenges to seabird research has beenuntangling the differential impacts of environmentalvariation and anthropogenic activities on seabirdpopulation dynamics.

Seabirds are important indicators of the status andstructure of marine ecosystems, because as wide-spread organisms, they inhabit all ocean systems(from inshore to pelagic, polar to tropical) and arerelatively easy to monitor at large spatio-temporalscales compared to other upper-trophic marineorganisms, (Furness & Camphuysen 1997, Durant etal. 2009). By nature of their reliance on the marineenvironment for food, and on terrestrial habitats forbreeding, research on seabirds provides a ‘window’into both marine and terrestrial systems (Grémillet &Charmantier 2010).

While ongoing research informs much of seabirdmanagement strategies, identifying the most criticalknowledge gaps in seabird ecology that limit effectiveseabird management and conservation is a timely andimportant exercise. Following the initiatives of Suther-land et al. (2006, 2009) and Hamann et al. (2010) that

identified key ecological questions relevant to biodi-versity conservation and policy, a group of researchersworking in seabird biology, ecology and conservationcompiled a list of 20 high-priority research questions.The present paper reflects the expertise of a group ofseabird researchers with a broad range of back-grounds, and details the top priority conservation-re-lated research questions for seabirds. Through thisprocess, we aim to strengthen the role of seabird re-search in facilitating and improving the conservationstatus of seabirds. Although this exercise focused par-ticularly on seabirds, the threats and challenges toseabird con servation are relevant to many othermarine top predators, such as sharks, tunas, marinemammals and turtles (Heithaus et al. 2008).

METHODS

Using ISI Web of Science, the first 6 authors iden-tified 180 candidate participants based on the num-ber of publications between 2006 to 2009 for‘(seabird or marine bird) and conservation’ (accessed10 April 2009). From this list we selected 35 partici-pants (30 male, 5 female) who represented the mostcomprehensive geographic expertise across oceanbasins as well as technical subject expertise. Eachparticipant was asked to list the 10 most important‘research questions that would assist in effectiveseabird con servation over the next decade’. Twenty-nine researchers (25 male, 4 female) from 9 coun-tries responded to the questionnaire (see Supple-ment 2 at www.int-res.com/articles/suppl/n017p093_supp. pdf. Responses were anonymous, with thecompilations only shared with the wider group onceall questions had been submitted.

The resulting 242 research questions were groupedinto broad categories by the first 6 authors, who cre-ated a short-list of research questions that wasshared with all participants. The list of the top 20research priorities was developed through partici-pant discussions, based on the number of responses,but does not reflect a relative ranking among priori-ties. Once the list was finalized, small groups of par-ticipants selected one of the broad research topicsand worked collaboratively to develop the support-ing text. All authors reviewed the entire manuscript.

RESULTS

We identified 6 broad research topics that encom-passed the specific research priorities identified by

94

http://www.int-res.com/articles/suppl/n017p093_supp.pdfhttp://www.int-res.com/articles/suppl/n017p093_supp.pdf

Lewison et al.: Research priorities for seabirds

the participants. These topics represent the funda-mental elements of seabird ecology, environmentalchange and cumulative impacts of human activity onseabirds and their ocean ecosystem.1. Population dynamics1.1. What are the key factors that regulate seabird

popu lations?1.2. How prevalent and influential are demographic

anomalies or cycles, e.g. breeding failures, peri-ods of decreased survival, disease outbreaks?

1.3. How can we best quantify demographic para-meters, popu lation size, population trends andrisks of extinction per species and per colony?

1.4. What is the form and importance of populationstructure?

1.5. How do the interactions among seabird speciesinfluence seabird populations?

2. At-sea spatial ecology2.1. How can we address knowledge gaps in at-sea

distribution?2.2. What habitat features delineate movement cor-

ridors or residency areas and how can re -searchers identify these corridors and bestquantify movement behavior?

3. Trophic dynamics and community roles of sea -birds

3.1. What are the roles of seabirds in communitiesand food webs?

3.2. How can we define, identify and map key forag-ing areas?

3.3. How do fisheries-mediated changes in trophicstructure influence seabirds?

4. Direct effects of fisheries4.1. How can we obtain detailed, unbiased and

accurate measures of bycatch rates?4.2. How do we evaluate bycatch risk to different

species or colonies in space and time?4.3. What are the population-level effects of by -

catch?4.4. What are the positive and negative effects of

discards in provisioning seabirds?5. Global change and population response to en -

vironmental variability5.1. How resilient are seabirds to climate and related

environmental change?5.2. What are the likely cascading trophic effects

on seabird population from environmentalchange?

5.3. How do interactions at different hierarchicalscales affect seabirds’ responses to environmen-tal change?

6. Managing anthropogenic impacts (invasive spe-cies, contaminants and protected areas)

6.1. What are the population-level impacts of inva-sive species?

6.2. What is the population-level influence of conta-minants and other pollutants?

6.3. How effective are protected areas in protectingseabirds?

1. Population dynamics

Understanding the factors that regulate and influ -ence populations is an essential component of sea birdconservation. Demographics are key variables in as-sessing extinction risk for populations and species(IUCN Red List criteria, Mace et al. 2008). Demo-graphic rates drive population trends, and accurateestimation of demographic parameters facilitates thediagnosis of environmental causes of populationchange (Green 1995) and provides a framework forpredicting responses to management (Frederiksen etal. 2004, Wanless et al. 2009). In general, seabirdshave high survival rates, low fertility and a progressiverecruitment with age, with an immature phase thatcan last for several years. They mostly breed incolonies and are philopatric to the natal site, althoughdispersal (both natal and breeding) occurs. Forseabirds, estimation of underlying demographic ratesis particularly important because declines in breedingpopulations can be influenced substantially by non-breeders and immature birds (Grimm et al. 2005,Votier et al. 2008a). The influence of non-breeders andimmature birds is particularly strong when populationsare density-dependent and large numbers of sexuallymature birds are queuing for recruitment; these birdscan rapidly integrate into the breeding populationwhen additive mortality occurs (Tavecchia et al. 2007,Votier et al. 2008a). The capacity of nonbreeders tomask, mitigate or exacerbate the effect of ecologicalperturbations (both human and natural) is poorly un-derstood because individuals cannot be easily moni-tored once outside the breeding site (Oro et al. 2006b).

Changes in population size influence the persis-tence of populations, and high levels of fluctuation,particularly in small populations, are associated withan increased probability of extinction (Engen et al2005). Seabirds are long-lived organisms and, assuch, their population dynamics are most immedi-ately sensitive to changes in adult survival, althoughchanges in reproductive success and juvenile sur-vival are important over time (Weimerskirch 2002,Finkelstein et al. 2010). Annual census informationon the number of nesting individuals, age at firstbreeding, and mean fecundity are widely available

95


for many seabird populations, e.g. for the ca. 275 spe-cies of seabirds (including Sphenisciformes, Procel-lariformes, Pelecaniformes and some Charadri-formes); information on age at first breeding isavailable for more than 60% of the species. Howeverother basic demographic information, including sur-vival of younger age classes, percentage of non-breeding birds in the population, fecundity followingmate loss, senescence, and the strength of natal andbreeding site fidelity are still lacking for manyseabird species. The percentage of species for whichadequate data exist decreases for these parameters:ca. 20% of species have estimates of adult survival,less than 5% of species have a known recruitmentcurve, and fewer than 2% of seabird species haveestimates of juvenile survival. These data gaps mustbe addressed if we are to understand seabird popula-tion response to episodic environmental events andthe capacity of seabird populations to recover fromhuman-mediated and other environmental changes.

1.1. What are the key factors that regulate seabird populations?

Seabird populations fluctuate as a result of complexinteractions between environmental characteristicsand vital rates. Vital rates (e.g. survival, fecundity, re-cruitment, dispersal) can also vary as a result of de-terministic factors (many of them related to humanactivities, e.g. human-mediated invasions) and alsoby the effects of density-dependence, both negative(e.g. Ruiz et al. 1998, Tavecchia et al. 2007) and posi-tive (i.e. Allee effects, Oro et al. 2006b). As such, all ofthese variables act as key factors regulating seabirdpopulations. Even though many of these factors havebeen analyzed independently in the literature, theirrelative importance is poorly documented and likelychanges over time. As with most species, seabirdpopulations are directly affected by predation (Gil -christ 1999, Oro et al. 1999, Bonnaud et al. 2009), aswell as by food and habitat availability. Environmen-tal fluctuations can indirectly drive all these factors(Doherty et al. 2004, Bertram et al. 2005, Oro et al.2010). Seabirds have evolved a life history strategythat relies on iteroparity — repeated fecundity — overa relatively long lifetime, and seabird populationgrowth is typically most sensitive to changes that re-duce adult survival. However, the importance ofchanges in productivity, including complete breedingfailures, can impact populations as well. Recent stud-ies have shown that breeding success andrecruitment have a strong impact on fluctuations in

some populations (Jenouvrier et al. 2005a,b, Ezard etal. 2006). Dispersal, undertaken by immatures andadults as natal and breeding dispersal, respectively,is also a key parameter in seabird population dynam-ics as it can greatly influence population growth ratesand extinction risks (Spendelow et al. 1995, Inchausti& Weimerskirch 2002, Cam et al. 2004). The metapop-ulation paradigm, together with the proper estimationof dispersal rates (see Section 1.4), highlights the dis-crete distribution of colonies in space and the vulner-ability of isolated populations with limited connectiv-ity (Cam et al. 2004, Genovart et al. 2007).

1.2. How prevalent and influential are demographic anomalies or cycles, e.g. breeding failures, periods

of decreased survival, disease outbreaks?

Abiotic environmental perturbations leading topopulation anomalies in seabirds include long-termand short-term oceanographic shifts, such as theNorthern Atlantic Oscillation (NAO, e.g. Durant etal. 2004), El Niño-Southern Oscillation (ENSO;Schreiber & Schreiber 1984, Veit & Montevecchi2006), the Pacific Decadal Oscillation (PDO) (Mantuaet al. 1997) or the Indian Ocean Dipole (Cai et al.2009) changes in the Antarctic Polar Frontal Zone,and also long periods of onshore atypical winds,tsunamis and storms (Wilson 1991, Harris & Wanless1996, Olsson & van der Jeugd 2002, Barbraud &Weimerskirch 2003, Frederiksen et al. 2008, Devneyet al. 2009). These events can affect multiple demo-graphic parameters at different time scales. The mostimmediate response is likely to be detected in annualbreeding success because, as long-lived organisms,seabirds skip breeding or abandon the nest if theirsurvival is jeopardized (Tavecchia et al. 2007, LeBohec et al. 2008). Adult survival, even though it isthe life-history trait most buffered against environ-mental stochasticity, may also decrease as a result offood shortages and reduced reproductive effort (Har-ris & Wanless 1996, Olsson & van der Jeugd 2002,Sanz-Aguilar et al. 2008). Even temporally limitedevents, i.e. storms, can have dramatic effects on pop-ulation dynamics. Harris & Wanless (1996) estimatedthat a shag population, Phalacrocorax aristotelis,affected by a winter storm would take 10 yr torecover from a population crash which took it to itslowest level in 35 yr. ENSO, a recurring multi-yearregime shift, has been linked to more marked popu-lation responses, although many of the observeddeclines reverse when ocean conditions revert (Mur-phy 1923, Wilson 1991, Barbraud & Weimerskirch

96


2003). Long-term, decadal ocean trends have beenlinked to crashes in prey availability (Devney et al.2009), which have resulted in reduced adult survival,a sudden drop in breeding frequency, or completebreeding failures (Montevecchi & Myers 1995).Ocean changes triggered by large-scale oceanicphenomena such as the ENSO and the NAO havebeen linked to severe negative impacts on both tem-perate and tropical seabird populations (Schreiber &Schreiber 1984, Bertram et al. 2005, Oro et al. 2010).

Disease outbreaks represent a biotic type of demo-graphic anomaly. Although seabirds seem to havevery competent immune systems (Esparza et al.2004) and live in vector-poor environments, emer-gent infectious diseases can cause chick mortality,with apparent cyclic patterns among years, reducingpopulation growth rates (Weimerskirch 2004). Syner-gistic effects between changing environmental con-ditions and disease should be explored further (Rol-land et al. 2009), especially as little is known aboutfactors affecting the circulation of infectious agents atdifferent spatial and temporal scales.

1.3. How can we best quantify demographic parameters, population status and risks of extinction

per species and per colony?

As discussed, for the majority of the world’s seabirdpopulations, demographic information is lacking. Forboth extremely large colonies, such as some SouthernOcean and northern ocean petrels) and for rare spe-cies in urgent need of conservation action, thenumber of individuals in a colony can lead to chal-lenges in obtaining demographic data. There are sub-stantial logistical obstacles to describing the demog-raphy of tropical seabirds with protracted breedingseasons (Ratcliffe et al. 2008), and with analyzing thesurvival and recruitment of immature birds given (of-ten unknown) immigration and emigration patternsand population network structure, e.g. source-sinks.This results in a bias towards data collection on tem-perate/polar populations, and adult birds.

Recent progress in analytical approaches has led toadvances despite these challenges. Distance-sam-pling, its extensions into capture-mark-recapture(CMR), spatial modeling designs, and other recent re-finements (Marques et al. 2007, Buckland et al. 2008,Thomas et al. 2010) are improving estimations of pop-ulation size and trend (Barbraud et al. 2009, Southwell& Low 2009), and streamlining fieldwork programs(Kendall et al. 2009). The use of meta-analysis andsurrogate taxa offers potential for rapid estimation of

key demographic parameters in poorly studied popu-lations (Brooke et al. 2010). Ratcliffe et al. (2008)demonstrated a promising approach, termed ‘virtualseabird’ modeling, to estimating population size in atropical seabird with protracted breeding. The rapiddevelopment of data modeling techniques that extendCMR ideas into multi-state models (White et al. 2006,Converse et al. 2009, Lebreton et al. 2009) and space-state models (Gimenez et al. 2008, Patterson et al.2008) opens new avenues for unbiased estimation ofdemographic variables, including immature survival/ -recruitment (Jenouvrier et al. 2008, Votier et al.2008a, Tavecchia et al. 2009). Likewise, genetic ap-proaches relevant at the spatiotemporal scale of popu-lation demography are now being deployed to esti-mate immigration/emigration rates (Milot et al. 2008,Peery et al. 2008, Boessenkool et al. 2009). Genetictechniques are also being employed to census data-poor populations using a CMR approach (Lawrence etal. 2008). Finally, electronic techniques for remote,automated and passive data-gathering are rapidly de-veloping (Wilson et al. 2002, Adams & Flora 2009).

1.4. What is the form and importance of populationstructure?

Most seabird species are spatially structured in dis-crete, patchy breeding sites or colonies. Although thehighly vagile nature of seabirds would suggest highconnectivity and thus little genetic differentiationamong subpopulations, other common features inseabirds, such as strong site fidelity, or historical bar-riers have led to substantial genetic structure in some(Abbott & Double 2003, Dearborn et al. 2003, Friesenet al. 2006), but not all, seabird populations (Burson1990, Avise et al. 2000, Genovart et al. 2003, McCoyet al. 2005a, Riffaut et al. 2005, van Bekkum et al.2006, Schlosser et al. 2009). Many factors, both bioticand abiotic, may shape population structure inseabirds (Friesen et al. 2007), and it is these factorsthat need to be examined within and among popula-tion networks.

Despite a strong conceptual understanding of spa-tial structure, the importance of seabird populationstructure is unresolved for many populations. Popu-lation structure is influenced by differences in habi-tat quality between source and sink areas, which typ-ically reflect high and low quality sites, respectively.Although seabirds assess habitat quality of breedingareas (Danchin et al. 1998), some apparent high-quality areas have been shown to act as populationsinks (Ainley et al. 1990). Additionally, some studies

97


have shown that adjacent colonies experiencing sim-ilar environmental conditions can exhibit differentdemographic responses (Tavecchia et al. 2008, Sanz-Aguilar et al. 2009). Furthermore, even with knowl-edge of spatial structure, colonies are often studiedas a discrete unit and not as a part of a larger net-work. Ongoing research highlights the importance ofconsidering the hierarchical spatial structure atwhich processes are occurring. Behavioral factorsaffecting dispersal and recruitment among coloniesor sub-colonies are an important consideration,notably in relation to the natural and changing pat-terns of temporal variability of habitat suitability atdifferent spatial scales (Boulinier & Lemel 1996).

One of the central challenges to understandingpopulation structure in seabirds is being able to dif-ferentiate movement from gene flow. Movement pat-terns of banded birds may reveal a species’ dispersalcapabilities; however, movement per se does nottranslate into gene flow, i.e. effective dispersal. Cau-tion is also required when quantifying dispersalranges using marked individuals, because the aver-age dispersal distances can be underestimated andcan be episodic (Koenig et al. 1996). Additionally,many seabirds breed in remote or inaccessibleislands or cliffs so resighting data may be also biasedbecause of limited detectability. Molecular tools facil-itate the estimation of the estimation of effective dis-persal rates (Dieckmann et al. 1999, Hedrick 2001).

Population structure and connectivity amongcolonies likely plays an important role in the viabilityof seabird populations. Structure and connectivitycan increase overall population viability by eitherallowing recolonization of suitable but unoccupiedcolonies or re-colonizing locally extinct colonies(Levins 1970, Dieckmann et al. 1999, Hanski & Gag-giotti 2004). However, connectivity can also synchro-nize local population dynamics across subpopula-tions, and this has been found to increase thelikelihood of global extinction (Heino et al. 1997).Additionally, dispersal, once treated as a fixed trait, isnow considered a flexible trait that may change withtime and space depending on environmental andindividual conditions (Clobert et al. 2001, Bullock etal. 2002, Hanski & Gaggiotti 2004). Natural andhuman-mediated environmental shifts (e.g. habitatdestruction, fisheries mortality, fishery-induced foodweb changes, the introduction of invasive species,climate change) may affect connectivity amongcolonies and thus population viability. From a conser-vation perspective, there is a strong need to developpredictive models that include changes in connectiv-ity under different environmental conditions.

1.5. How do the interactions among seabird speciesinfluence seabird populations?

Many seabird species are sympatric with others,sharing breeding and foraging habitats. Potentialcompetition among species for prey was identifiedseveral decades ago (e.g. Ashmole 1968, Cody 1973)and has since been studied in tropical, polar and tem-perate ecosystems (Furness & Birkhead 1984, Fur-ness & Barrett 1985, Weimerskirch et al. 1986, Cairns1987, Ballance et al. 1997, Forero et al. 2004). Typi-cally, there is some degree of ecological segregationamong seabird species even if they exploit similarprey within a common location, for example in termsof prey sizes, foraging depths (Thaxter et al. 2010) ortime of foraging (Arcos et al. 2001, Spear et al. 2007).It is unclear, however, whether such segregation isdue to competitive exclusion or simply reflectsanatomical and behavioral adaptations to exploitingsubtly different prey (e.g. Whittam & Siegel-Causey1981). When sympatric species are similar in size,dietary overlap can be high and environmental sto-chasticity seems to affect them similarly (Bryant &Jones 1999). From a conservation perspective, over-harvesting of marine organisms by fisheries can notonly decrease prey availability for seabird communi-ties, but also alter the ecological relationships amongseabirds (Furness 2000, Montevecchi 2001, Votier etal. 2004). Generalist species, exploiting fishing dis-cards or dumps, have steadily increased, in somecases, to the detriment of specialist seabirds becauseof interference competition at breeding sites or den-sity-dependent competition for foraging resources.This is especially true when generalist species arelarge and aggressive, e.g. large gulls or skuas, andevidence suggests that human-induced changeshave already altered the structure of seabird commu-nities, although the broader ecological consequencesof these shifts among conspecifics are unclear (Ain-ley & Boekelheide 1990, Howes & Montevecchi 1993,García Borboroglu & Yorio 2007, Oro et al. 2009).Likewise, relatively little is known about the role ofcompetitive exclusion in breeding habitat selectionin seabird communities (Whittam & Siegel-Causey1981, Ainley & Boekelheide 1990). Conservationactions can alter nesting environments, favoring cer-tain species more than others within the community,and local extirpation of smaller seabirds by largerseabirds has been reported as a result of competitiveexclusion in suitable nesting habitat (Oro et al. 2009).

Although predatory interactions among seabirdspecies are less well studied than competitive inter-actions, they do occur (Gilchrist 1999, Votier et al.

98


2004, Oro et al. 2006a, Oro & Martínez-Abraín 2007).Large opportunistic species (e.g. gulls, giant petrels)can be facultative predators on smaller seabird spe-cies within the community, and population increasesof the former following the appearance of largeamounts of food from human activities may increasethe predation rates on vulnerable species (Regehr &Montevecchi 1997, Votier et al. 2004, Oro et al.2006a). Seabirds may also more or less share para-sites and pathogens that are prevalent in their breed-ing habitat, although the interactions involved arelikely complex. Host specialization has notably beensuggested for the common seabird tick Ixodes uriae(McCoy et al. 2005b), which can transmit severalviruses and bacteria. Mutualism may also play animportant role within some seabird communities,particularly in terms of prey detection (Jaquemet etal. 2005, Ainley et al. 2009). Some species are partic-ularly good indicators of fish location for otherseabird species, and changes in their numbers canhave consequences for the whole community (Catryet al. 2009).

2. At-sea spatial ecology

Research on at-sea distributions of seabirds hasserved to clarify the role of seabirds in marine com-munities, identify important ecological areas and topredict avian responses to environmental changes inthe marine environment. The central challenge is tounderstand how seabird spatial and temporal dynam-ics are linked to the spatial and temporal variability ofprey and related oceanographic features. The ‘classi-cal’ age of at-sea investigation of seabirds occurredduring the 1970–90s, when extensive data from ship-ping and other vessels were collected. During this pe-riod, there were major discoveries of the associationof seabirds and seabird communities with specificoceanographic features (Brown et al. 1975, Pockling-ton 1979, Briggs et al. 1987, Wahl et al. 1989), andfrontal zones at large spatial scales (Schneider et al.1987), as well as more fine-scale features (e.g. Hunt &Schneider 1987, Louzao et al. 2006) at smaller scales(Hunt & Harrison1990, Hunt 1991). The advances inremote sensing, increased computation resources,and major developments in tracking technology haveopened up even more opportunities to evaluateseabird associations with dynamic marine features,such that concurrent modeling of seabird and oceano-graphic data in a dynamic geospatial environment ispossible (Tew Kai et al. 2009, Tremblay et al. 2009,Wakefield et al. 2009).

2.1. How can we address knowledge gaps in at-sea distribution?

A lack of detailed knowledge about at-sea distribu-tion of individuals across species, sex and age classeshas presented one of the most substantial challengesto seabird ecology. Satellite and global location sensor(GLS) devices deployed post-breeding have providedinformation on sex-related specific movements foradults (e.g. González-Solís et al. 2000). GLS logginguses the timings of sunrise, sunset and the resultantday length and timing of local noon to derive an esti-mation of global position. Subadults, unlike adults,spend years at sea, which precludes instrument re-trieval, and as a result, little is known about their spa-tial distribution, except when plumage separation ofage classes is possible (as for gulls, penguins, frigate-birds; Ainley et al. 1984, Diamond & Schreiber 2002).Nevertheless, recent long-term deployment of GLSdevices on juveniles may shed some light on this issuein the near future. Alternatively, the use of intrinsicmarkers, such as stable isotopes, genetics and traceelements, can provide valuable information on thebroad-scale origin and movements of seabirds (Cherel& Hobson 2007, Gómez-Díaz & González-Solís 2007).Although data from intrinsic markers can provideonly coarse-scale information, these methods repre-sent a robust approach to mapping the spatial distrib-ution of immatures, non-breeders or species too smallfor instrumentation. Information on at-sea distribu-tions of seabirds has important applications. Thesedata have been instrumental in defining ImportantBird Areas (IBAs) and assessing potential interactionswith anthropogenic factors such as some fisheries,shipping, oil spills or off-shore wind farms (Louzao etal. 2009). Aggregated, these data have been used topresent the most comprehensive perspective on theglobal nature of seabird species (BirdLife International2004, Halpin et al. 2009).

Remotely sensed oceanographic traits data andtracking technology also increase our understanding ofseabird ecology and have shed new light on howseabirds associate with productivity gradients and bio-physical discontinuities at several spatiotemporal scales(Ballance et al. 1997, Spear et al. 2001, Hyrenbach et al.2006, Bost et al. 2009). In addition to distribution andmovement, tracking devices can also be used to inferfeeding activity of birds through the analysis of highresolution global positioning system (GPS) data, partic-ularly when these data are combined with other de-vices, such as depth and activity recorders or stomachtemperature loggers (Bost et al. 2008). Moreover, de-ployment of miniaturized sensors on diving seabirds

99

can also provide data on hydrographic conditions, suchas temperature and salinity, in foraging areas (Wilsonet al 2002, Bost et al. 2008, Durant et al. 2009).

Further progress towards understanding seabirdmarine ecology will strongly depend on real-timeinvestigations of ocean dynamics, both biological andphysical. A sound theoretical framework consideringa hierarchy of spatial scales has been developed toadvance this research (Fauchald 2009). At-sea sur-veys combined with intensive synoptic multi-discipli-nary investigations of ocean processes and food webdynamics, as well as tracking devices, can providecomprehensive information about at-sea behavior,and about what factors influence seabird distributionat the population and the individual level (Chapmanet al. 2004, Ribic et al. 2008, Ainley et al. 2009). Pro-jects such as GLOBEC exemplify this multi-discipli-nary approach (www.globec.org).

2.2. What habitat features delineate movement corridorsor residency areas and how can researchers identify

these corridors and quantify movement behavior?

Seabird species vary in the extent and directionalityof their movement patterns from within-region move-ment to long-distance migration. Networks of ob-servers and radars have provided valuable multi-spe-cies information on seabird movements and timing ofmigration, although much of this information hasbeen limited to coastal areas and specific vantagepoints (Mateos & Arroyo 2010). At-sea investigationsof seabird flight behavior can reveal movement pat-terns with sufficient environmental data (Spear & Ain-ley 1999, Shaffer et al. 2006). Defining movement cor-ridors between breeding colonies and foraginggrounds or among breeding, staging and winteringareas depends on the combined use of several track-ing systems. Integrating concurrent information fromtracked individuals and remotely sensed ocean condi-tions can be used to identify broad-scale seabirdmovement (González-Solís & Shaffer 2009, González-Solís et al. 2009, Adams & Flora 2010). The ongoingchallenge to this research is the need to link high-res-olution seabird location data with real-time data onphysical and biological oceanographic conditions.

3. Trophic dynamics and community roles ofseabirds

As top predators, marine birds are influenced byfood web dynamics that can be changed and per-

turbed by both top-down and bottom-up influencesas well as by physical oceanographic shifts (e.g. Fred-eriksen et al. 2007). Anthropogenic influences associ-ated with fisheries can induce pervasive top-downand indirect effects on food webs (e.g. Frank et al.2005) and hence on apex avian predators. Seabirddata are often used as a signal of shifts in marine foodwebs (Piatt et al. 2007). Synoptic vessel surveys,tracking studies and stable isotope sampling are pro-viding comprehensive new information about the dis-tributions, movements, activities, physical environ-mental measurements and foraging sites of seabirdsthroughout the year. Analytical methods for delineat-ing core habitat areas are being integrated and re-fined. Risk analyses of anthropogenic factors associ-ated with fishing, oil pollution, and shipping can beused to identify problem areas and to help resolveconflicts with adaptive management strategies.

3.1. What are the roles of seabirds in communitiesand food webs?

Understanding the role that seabirds play in com-munities and food webs is complicated by fluctua-tions in the ocean environment, the great distancescovered by the seabirds and the associated difficultyof observing interactions of seabirds and their prey(Croxall et al. 1999). Consequently, the functionalrelationships between seabirds and their biotic andabiotic environments remain elusive (Grémillet et al.2008). The technology to quantify habitat use by indi-vidual seabirds now exists, but quantitative analysesat appropriate scales are only just being used inhypothesis testing that could improve our under-standing of the processes linking seabird distributionto their environment (Tew Kai et al 2009, Tremblay etal. 2009, Wakefield et al. 2009).

The marine environment is also subject to a varietyof anthropogenic and natural stresses that affectmarine productivity, stochasticity and cyclicity (Hal -pern et al. 2008, Grémillet & Boulinier 2009). Conse-quently, marine food webs are often unstable andprone to regime shifts and numerous bottom-up andtop-down controls so that any one event may havemultiple causes and consequences. There is evidenceof trophic cascades resulting from overfishing and theremoval of large mammals and large predatory fishesin many open-ocean systems (Pauly et al. 1998, Franket al. 2005, Baum & Worm 2009). In the Ross Sea,where an immense upper trophic level community isbuilt upon the high primary production of this region,Ainley et al. (2004, 2006, 2007) showed that large



numbers of competing top predators, including pen-guins, flighted birds, seals and whales, can decreasethe available density of krill and fish prey species,thus reducing the grazing pressure on phytoplankton.As a result, a significant portion of phytoplankton re-mains ungrazed and sinks to the benthos, thus enrich-ing that food web (Ainley et al. 2007).

The degree to which seabirds serve as indicators oftrophic shifts within marine ecosystems varies withthe species of seabird and their degree of prey spe-cialization and the variability of their foraging behav-ior (e.g. Montevecchi & Myers 1995, Croxall et al.1999, Hamer et al. 2007). Seabird species that aredependent upon a limited array of prey species andmay have less varied feeding methods are likely toexhibit more significant inter-annual variations inbreeding frequency and success than those seabirdswhich feed on a wide variety of prey taken by a vari-ety of feeding methods (Berrow & Croxall 1999).However, trophic shifts triggered by intense large-scale oceanic phenomena such as the ENSO and theNAO have been linked to severe negative impacts onboth temperate and tropical seabird populations(Schreiber & Schreiber 1984, Bertram et al. 2005; seeSection 1.2). In general, these impacts are attributedto significant changes in productivity at lower trophiclevels, (Devney et al. 2009) which result in reducedadult survival, a sudden drop in breeding frequencyand/or complete breeding failure. The robustness ofthese studies indicates that, aside from catastrophicimpacts of large-scale oceanic phenomena, studies ofprey species composition, levels of prey harvest, andforaging duration of seabirds can provide reliableindices of prey abundance and thus serve as indica-tors of trophic shifts (Montevecchi & Myers 1995,Croxall et al. 1999).

3.2. How can we define, identify and map keyforaging grounds?

As discussed in Section 2.1, the earliest studies ofat-sea distributions of seabirds were derived fromship-based surveys carried out opportunistically oron dedicated cruises (Harrison 1982, Haney 1985).Subsequent technological innovations have dramati-cally improved the capacity to track movements ofindividuals (Wilson et al. 2002, Phillips et al. 2007,Burger & Shaffer 2008). Use of satellite-transmittersand, in recent years, GPS loggers have thus providedextensive, high-resolution distribution data frommany medium to large seabirds during the breedingseason. Increasing use of GLS technology has pro-

vided low-resolution but, for some species, year-round data (Weimerskirch et al. 2002, Phillips et al.2005, Rayner et al. 2008, Egevang et al 2010). Com-plementing the telemetry innovations, recent track-ing data analyses have highlighted the importance offronts, upwelling, eddies and other recurring oceanfeatures that are indicators of high quality habitatwith abundant sources of prey (Bost et al. 2009). Withincreasing awareness of natural and anthropogenicimpacts on seabirds (e.g. the Convention on Biologi-cal Diversity [CBD] call to identify ecologically orbiologically significant areas [EBSAs] in the highseas), more effort has been focused on mapping andcharacterizing key foraging grounds and resources.

A number of challenges have limited the ability toidentify foraging areas. The first challenge is method-ological: What analyses should be used to determine aforaging area? A suite of alternative approaches havebeen proposed that include kernel density estimates,track sinuosity, First Passage Time analysis and state-space models (Wakefield et al. 2009). As with other re-search questions, incorporating real-time remotelysensed environmental information (with appropriatemeasurement error) is an essential element to developimproved frameworks for integrating vessel-basedand telemetric data, each of which suffers fromknown constraints (Louzao et al. 2009). To do this re-quires a range of device and data types, including im-mersion loggers, time-depth recorders, stomach tem-perature probes, accelerometers and multi-channelloggers (Weimerskirch & Wilson 1992, Wilson et al.2008), and experimental studies of the importance ofolfaction (Nevitt & Bonadonna 2005) in order to incor-porate information on fine-scale behavior and sensoryperception into feeding site identification. In addition,data on habitat accessibility and intra- and inter-spe-cific competition (Grémillet et al. 2004) are needed totrack habitat use as well as habitat preferences, andthus improve the capacity to predict spatial distribu-tion of birds.

Identification of key prey resources is another com-ponent to understanding foraging requirements ofseabirds (Pichegru et al. 2010). Conventionalapproaches to diet determination, which includevisual observations of birds returning to colonieswith prey in their bill, and sampling of stomach con-tents, pellets etc., are often effective during chick-rearing, albeit with some inherent biases (Barrett etal. 2007). Understanding prey preferences outside ofthe breeding season is more challenging. With suffi-cient tissue samples, recent developments in theanalyses of bulk and compound-specific stable iso-topes, fatty acids, and other signature elements and

101


compounds offer unprecedented opportunities tocharacterize diet in high taxonomic detail (Cherel etal. 2005, Connan et al. 2007, Käkelä et al. 2007,2010). Seabirds offer high potential as indicators ofbroad-scale environmental change and of processesoccurring at lower trophic levels, and therefore, byimproving identification and characterization of theirkey areas, habitats and resources, seabird ecologistscan make a vital contribution to current global initia-tives to improve protection and sustainable manage-ment of the oceans.

3.3. How do fisheries-mediated changes in trophicstructure influence seabirds?

Fisheries can change, and have clearly changed,and at times greatly perturbed marine food webs(Jackson et al. 2001, Hilborn et al. 2003, Estes et al.2006). The effects of fisheries on marine birds aredetermined by the fisheries’ target species, as well asthe intensity, persistence and the spatial distributionof fishing effort. Human-induced changes in trophicstructure can affect marine birds either negatively orpositively (Montevecchi 2001). Fisheries (a top-downforce) reduce the prey base and indirectly affectmarine birds through prey depletion, a negativeeffect. However, fisheries may also reduce the num-ber of competitors that rely on the same prey asseabirds, and this may yield a positive effect. Theseanthropogenic effects can interact with, and beamplified by, bottom-up ocean climate changes(Osterblom et al. 2007). Overfishing of large fishescan lead to increases in the abundances of smallerfishes. This is often evident when fisheries targetlarge predators that can include pelagic species(Sherman et al. 1981), demersal species (Bundy 2005)or large marine mammals (Springer et al. 2003, Esteset al. 2006). Because it is the indirect relationships infood webs that are often pervasive, complexity oftenoverrides straightforward rebounds of prey popula-tions once predators are removed (Yodzis 2001).Fisheries can also have direct positive effects by pro-viding food in the form of discards and offal thatwould otherwise be inaccessible to marine birds, yetthe positive effects of discards vary across and withintaxa (Garthe et al. 1996; see Section 4.4).

4. Direct effects of fisheries

As noted in Section 1, adult mortality has themost immediate effect on the population trends of

seabirds. Incidental capture in fishing gear, calledbycatch, is one factor that causes adult mortalityand has been the focus of research and conserva-tion concern since the late 1980s (e.g. Weimerskirch& Jouventin 1987, Bartle 1991, Brothers 1991).Longline, trawl, and gillnet fishing gear can alllead to bird injury and mortality as birds forage forbait fish and discards. Different fishing gear typespose different threats to seabirds (Phillips et al.2010). Birds scavenging or targeting baited longlinehooks may get hooked and dragged under whilethe gear is at the surface. Trawl gear causes mor-tality primarily by birds striking cables. Divingbirds may get entangled in gillnets while scaveng-ing fish caught in the nets or pursuing foragefishes. Discarded fish parts, commonly called offal,may also provision some scavenging species, pro-viding easily captured prey. Many researchers,agencies and governments have worked closelywith industry to develop a wide range of seabirdbycatch mitigation devices (Cherel et al 1996,Brothers et al. 1999, Melvin et al. 1999, 2001, 2004,O’Toole & Molloy 2000, Boggs 2001, Robertson etal. 2003, Gilman et al. 2005). When used together,these devices have been shown to reduce seabirdbycatch by more than 90% in some fisheries (Coxet al. 2007). Despite the increased awareness of theinfluence fisheries can have on seabird species andsuccessful bycatch reduction by some fleets, basicquestions remain on the extent of the negative andpositive effects of fisheries on seabird populations.

4.1. How can we obtain detailed, unbiased andaccurate measures of bycatch rates?

One of the central challenges to understanding thelong-term effects of fisheries bycatch on seabird pop-ulations is the lack of detailed data on bycatch ratesfor individuals, species (versus taxa), colonies, aswell as for age and sex classes within colonies(Gómez-Díaz & González-Solís 2007). Given theimportance of population sub-structure and move-ment for many seabird species (see Section1.4),understanding differential impact of fisheriesbycatch among colonies is an essential component tounderstanding the magnitude and spatial extent ofbycatch effects. Likewise, although bycatch leads todirect declines in survival rates, it may also havestrong effects on subsequent fecundity and sex ratios(Mills and Ryan 2005).

Obtaining accurate estimates of bycatch rates perspecies, colony and age or sex class requires direct

102


measurement from dead individuals, but also propersampling design and statistical treatment (e.g. Laneriet al. 2010). Carcass collection, identification andanalysis are instrumental in identifying species, sex-ing and aging the birds, to the highest resolution pos-sible. With carcasses in hand, samples can also becollected for subsequent genetic analyses to furtherclassify individuals to specific colonies, at least forspecies with clear population structure (Edwards etal. 2001, Gómez-Díaz & González-Solís 2007). Exist-ing carcass recovery programs point to sex-biasedmortality in some regions. Bycatch research aroundthe Prince Edward Islands found a strong adult malebias in bycatch mortality (Nel et al. 2002, Ryan &Boix-Hinzen (1999) for 3 species (white-chinnedpetrel Procellaria aequinoctialis, grey-headed alba-tross Thalassarche chrysostoma and Indian yellow-nosed albatross T. carteri), whereas studies aroundNew Zealand, South Georgia and the Crozet Islandshave found adult female-biased bycatch mortality forgrey petrels Procellaria cinerea and wandering alba-trosses Diomedea exulans (Weimerskirch & Jou-ventin 1987, Croxall & Prince 1990, Murray et al.1993). In some areas, the bias in capture is relative toage of the individual; juvenile seabirds may be dis-proportionately affected by gear interactions (Bakeret al 2007). Genetic and post-mortem analyses ofalbatrosses carcasses from longline fishing gear in 2New Zealand fishing areas illustrate the power of thisapproach to identify the age of individual, speciescaught, sex and colony of origin (Burg 2007).

The lack of widespread carcass-based workreflects, for the most part, the challenges involved inretaining large carcasses on space-limited fishingvessels. As with other analyses that consider theextent of mortality from a particular source, i.e.bycatch, oil contamination or storm events, there arealso limitations from direct carcass measurementsassociated with factors that can lead to biases in car-cass recovery, e.g. undetected carcasses that drop offor out of fishing gear (Ryan & Watkins 2008), carcassdegradation during gear deployments (Burg 2007),and changes in currents and ocean patterns in thecase of beached carcasses (Lucas & MacGregor 2006,Parrish et al. 2007). As with basic demographic para-meter estimation, models provide one means ofaddressing these biases and improving estimates oftotal mortality (sensu Flint et al. 1999, Wiese &Robertson 2004, Martínez-Abraín et al 2006, Veran etal. 2007). However, given the current data used,these models have limited ability to provide the highresolution estimates of mortality per colony, age orsex class.

4.2. How do we evaluate bycatch risk to differentspecies or colonies in space and time?

A variety of methods have been used to evaluate therisk of fisheries bycatch to seabirds of differentspecies, colonies, sexes and age classes (Piatt et al.1984, Prince et al. 1992, Brothers et al. 1997, Weimer-skirch et al. 2000). Traditionally, these approacheshave included surveys of seabirds attending vessels byobservers who record data on species and age classes,observations at colonies of information from bandedand color-marked individuals, and wildlife telemetrywhich records individual-level data of tagged individ-uals. Increasingly, new approaches are being devel-oped to search for evidence of fishery-related subsidiesin seabird populations (e.g. stable isotopes; Votier etal. 2010) and to identify the provenance of bycaughtindividuals using genetics, morphometrics, or stableisotopes (Gómez-Díaz & González-Solís 2007).

Following the fisheries management arena, theevaluation of bycatch risk can focus on the assess-ment of ‘inputs’ and ‘outputs’ into the system. Inputvariables quantify the risk to a given species or pop-ulation component by quantifying its spatio-temporaloverlap with fisheries effort. This approachhas 3 lim-itations: (1) it requires high-resolution concurrentdata on seabird dispersion and fishing effort; (2) itrelies on the underlying assumption that overlap isproportional to bycatch risk; and (3) it yields scale-dependent risk metrics (for example, the degree ofoverlap and the statistical association betweenseabirds and fisheries’ distributions vary with thetemporal and spatial scales of analysis). Thus, empir-ical studies comparing these risk factors with actualbycatch rates would provide valuable insights intothese underlying assumptions.

Output approaches quantify the effect of bycatch rel-ative to population structure, e.g. age class, sex, andthe geographic provenance (e.g. colony of origin) ofbycaught individuals. While these data can, in theory,provide an absolute metric of bycatch risk of a singlevessel, by comparing the contributions of differentpopulation segments to the observed bycatch, theycannot provide a comprehensive assessment of bycatchrisk across all vessels within or among fleets. Unfortu-nately, the lack of fishery-independent data on seabirdabundance and the differential reliance on fisheriesamong colonies and among age/sex classes limit theability to scale the observed impacts across all vessels.

Integrated approaches capable of merging theseinput and output perspectives are critical to developstandardized bycatch risk assessments. To this end,studies should integrate input metrics (e.g. surveys of

103


seabird distributions within fishing grounds, obser-vations of vessel attendance) with output metrics(e.g. bycatch rates by sex- and age-class) from thesame fisheries. In addition to determining the sexand age of bycaught specimens with necropsies,novel tracers are used to determine the provenanceof bycaught individuals (Gómez-Díaz & González-Solís 2007). Efforts to develop an integrated pictureof bycatch risk for specific seabird population compo-nents will benefit from the implementation of carcassretention and necropsy programs, and from thedevelopment of signature profiles for individualcolonies. The final step of this integrated assessmententails ground-truthing population models with thebycatch composition data from the field (Moloney etal. 1994, Tuck et al. 2001).

New techniques and comprehensive applicationsare needed to effectively study the effects of fisherieson seabirds and other large marine predators. Inte-gration of vessel surveys of birds at sea with mappinganalyses derived from animal-borne tracking deviceson free-ranging seabirds (Louzao et al. 2009) willprovide the best means of understanding avianoceanic distributions and occupancy of sites. Thesespace/time distributions have to be integrated withglobal databases and maps of fishing activities andcatches to better identify ocean biological hotspots(Roberts et al. 2002) and to execute risk analysesinvolving areas of overlap, interaction and potentialconflict (Karpouzi et al. 2007). These analyses need tobe pursued on global, ocean-basin and regionalscales (Bartumeus et al. 2010).

4.3. What are the population-level effects ofbycatch?

As discussed in Section 1.1, the population-levelimpacts of incidental mortality are influenced by thespecies-specific life-history characteristics, includingthe sensitivity of population growth rate (lambda) todemographic parameters (adult survivorship vs.reproductive success), and by the differential suscep-tibility of different age classes and sexes to this mor-tality (Russell 1999, Weimerskirch 2002). Thus, popu-lation models consistently conclude that fisheries thatpreferentially kill breeding adults are expected tohave a more drastic effect on lambda, since this ageclass has the highest survivorship rate and currentreproductive value is the most influential on lambda(Piatt et al. 1984, Moloney et al. 1994, Cousins 2001).

In turn, the susceptibility of different sexes and ageclasses to bycatch is expected to be influenced by dif-

ferential rates of interaction with fisheries. There ismounting evidence that different age classes (e.g.Suryan et al. 2007) and sexes (e.g. Nel et al. 2002)within a species differ in their overlap with fisheries,due to the segregation of at-sea distributions and dif-ferences in foraging ecology. While the influence ofthis ecological segregation on bycatch risks has beendocumented, especially for sexually dimorphic species,(e.g. Prince et al. 1992, González-Solís et al. 2007), thedegree to which different age classes rely on fisheriesis not well understood (but see Bartumeus et al. 2010,Votier et al. 2010). In particular, it is essential to knowwhether young birds are more susceptible to fishinggear, and whether this higher susceptibility is causedby their lack of experience or by a higher likelihood foryounger birds to forage at fishing vessels.

An improved assessment of population-level effectsof bycatch requires 3 ingredients. First, these impactsneed to be evaluated in a population-level context.This requires an understanding of the magnitude(number of individuals as a proportion of the total pop-ulation) and data on the differential risk to certain in-dividuals (i.e. birds that specialize on fisheries) or ageclasses (i.e. young of year). This information is criticalfor developing population metrics of bycatch riskusing data on spatial overlap (Prince et al. 1992) orvessel attendance (Weimerskirch et al. 2000). Second,it is imperative to place these species-specific bycatchrisks in a broader demographic context by balancingthe resulting bycatch mortality against the benefits offood subsidies, which may be critical food resourcesfor inexperienced birds during their first years at seaor during periods of low prey productivity. Third, be-cause the degree of vessel attendance and the benefitsversus costs associated with this behavior are likely in-fluenced by the presence of competitors, it is critical toinvestigate how seabird−fishery interactions, such asaccess to baited hooks and foraging success, are influ-enced by the presence and density of other species.Without this community-level understanding, bycatchmitigation measures attempting to minimize the takeof one protected species may cause unintended im-pacts on other species. For instance, night-time setsdesigned to avoid the competitively dominant large-bodied albatrosses, appear to increase the bycatch ofpetrels and shearwaters (Cherel et al. 1996).

4.4. What are the positive and negative effects ofdiscards in provisioning seabirds?

It is estimated that 7.3 million t of fish are discardedannually by marine fisheries throughout the world

104


(Kelleher 2004). Although this estimate is lower thanin previous decades (Alverson et al. 1994), when con-sidered in conjunction with associated overfishing,discards are likely to exert significant impacts onseabirds (Furness 2003). A common and overabun-dant food source, discards have artificially inflatedscavenging seabird populations (Furness et al 1992,Garthe et al 1996), probably due to improved bodycondition and breeding performance (Bunce et al.2002, Oro et al. 2004, Votier et al. 2008a). Docu-mented increases in northern fulmars in the NorthSea have been ascribed, in part, to food subsidiesfrom fisheries discards (Camphuysen & Garthe 1997,Thompson 2006).

Although discards may serve as an importantresource for some seabird populations, they havealso been documented to have a negative influenceon other seabird populations (Kitaysky et al. 2006,Pichegru et al. 2007, Österblom et al 2008, Grémilletet al. 2008, Navarro et al. 2009). The long-termimpacts of discards on seabirds, both positive andnegative are still emerging (Arcos et al. 2008, Bar-tumeus et al. 2010). The proportion of foraging indi-viduals or age classes relying on discards is unclear(but see Navarro et al. 2010), as is whether speciesare developing discard specialization. As discussedin Section 1.2, generalist species that exploit fishingdiscards have steadily increased, in some cases, tothe detriment of specialist seabirds. A combination ofdiet studies, biomarkers (e.g. fatty acids, stable iso-topes) and field studies is needed to effectively studyhow discards are influencing demographic parame-ters and population dynamics.

5. Global change and population response toenvironmental variability

Human-induced disruption of ocean resources hashad a profound effect on ocean food webs and com-munities, although the magnitude of these effects isnot easily detected (Pauly 1995, Pitcher 2001, Pauly &Maclean 2003). Even the carrying capacity of oncerich areas has been severely reduced (Christensen &Richardson 2008, Watermeyer et al. 2008a,b). In thiscontext, global-scale environmental changes, partic-ularly rising temperatures, retreating sea ice andocean acidification, are expected to be the dominantanthropogenic pressures on marine ecosystems inthe next decades (Harley et al. 2006). These effectsare exacerbated by the simplification of food websowing to fish depletion (Cury et al. 2000, Osterblomet al. 2007, Watermeyer et al. 2008a,b). Current and

projected changes are rapid and large, and condi-tions are likely to shift to entirely novel states(Solomon et al. 2007). Understanding how seabirdpopulations are affected by changing conditions istherefore critical to support the development of effi-cient conservation efforts. Research needed toadvance understanding of these issues includes inte-grated comparisons across study sites, eco-physio-logical studies, and collaborations with researchersworking on other trophic levels and on physical vari-ables (e.g. physical oceanographers, climatologists).

5.1. How resilient are seabirds to climate andrelated environmental change?

Meaningful conservation efforts for seabirds in the21st century need to take into account how resilientpopulations and species are to climatic and otheroceanographic changes. Will seabirds be able tochange their breeding and non-breeding ranges, for-aging habitats, diets etc. sufficiently quickly to keepup with the increasing pace of change (Grémillet &Boulinier 2009)? Can change through phenotypicplasticity facilitate adapation, or will evolutionaryprocesses make seabird survival possible? Thesequestions cannot profitably be addressed for allseabirds, because species differ widely in life historycharacteristics, exposure to other stressors and, thus,in resilience. A wide variety of research tools,applied at different organizational levels, will be nec-essary to provide answers for seabird species, com-munities and ecosystems.

At the individual level, we need to understand towhat extent seabirds can adapt to the direct, e.g.increasing temperatures, and indirect effects of envi-ronmental change from changes in prey availability.Ecophysiological studies play an important role, bothin an experimental, captive setting and for free-rang-ing birds in their natural environment (Gaston et al.2009). Empirical studies of energy balance, e.g.including experimental manipulation of food supply,can provide information on the limits of phenotypicplasticity in terms of adapting to changing food avail-ability and quality (Enstipp et al. 2007). Energeticsmodeling has also been demonstrated as a viablemeans of assessing the impact of environmentalchange on seabird energy balance (Fort et al. 2009).Likewise, studies of thermal tolerance will allow theconstruction of species-specific Scholander curves,characterising the lower and upper critical tempera-ture thresholds as well as the temperature comfortzone (Scholander et al. 1950).

105


In order to affect populations, impacts of environ-mental change on seabirds must manifest themselvesas changes in demographic parameters (reproduc-tion, survival and dispersal). Estimating realizedchanges in these parameters requires data fromlong-term studies, preferably of marked individuals(Barbraud & Weimerskirch 2003, Frederiksen et al.2004). Although such studies are by their nature cor-relational and cannot be used to draw strong conclu-sions about causal relationships, they allow inferenceabout processes on a temporal and spatial scale thatis unattainable for experimental studies. In particu-lar, combining data from several geographically sep-arated populations allows a degree of replication andhas the potential to provide information on a range-wide scale (Grosbois et al. 2009). Data on at-sea dis-tributions can be treated in a similar way, althoughthe specific statistical methods differ.

There is an increasing realization that evolutionaryprocesses commonly occur on an ecologically rele-vant time scale, and in order to assess how organismsadapt to directional environmental change, anunderstanding of selection pressures and associatedresponses is necessary. However, estimating heri-tability and separating phenotypic plasticity frommicroevolution in wild animals is challenging, partic-ularly because detectability (e.g. recapture orresighting probability) is generally less than one(Gimenez et al. 2008). The most promising approachis based on long-term studies with extensive pedi-gree information, but such studies are rare (Doligezet al. 2009, Ozgul et al. 2009).

To predict the consequences of environmentalchange on distribution, abundance or viability ofwhole species or communities, modeling tools areneeded (Watermeyer et al. 2008a,b). At the simplestand most phenomenological level, climate envelopeand similar models predict future breeding or non-breeding distributions based on current empiricalassociations between the presence of a particularspecies and various aspects of climate (Guisan &Thuiller 2005). Such models may be more suitablefor non-breeding distributions where detailed infor-mation about individuals is limited, particularly ifthey can be linked to models of the future distribu-tion of important prey organisms (see Section 5.2),or for species breeding on low-lying islands threat-ened by sea level rise. A more mechanisticapproach uses matrix-based population models,which are parameterized with the results of demo-graphic studies, including any observed relation-ships between demographic performance and cli-mate (Caswell 2001). Finally, the most complex

approach uses the results of studies of individualbehavior, diet, and energetics directly to predictthe fate of individuals constituting virtual popula-tions (Sutherland & Norris 2002, Fort et al. 2009).Such behavior-based (or agent-based) models arepotentially highly realistic, but also very difficultand resource-demanding to parameterise. Thechoice of modeling approach should depend on thespecific questions and the data at hand.

5.2. What are the likely cascading trophic effects onseabird population from environmental change?

Marine ecosystems are and will be stronglyaffected by increasing temperatures and acidifica-tion due to increasing atmospheric CO2 concentra-tions. Increasing temperatures lead to changes inocean currents, primary productivity (Behrenfeld etal. 2006), decreasing ice coverage in polar regionsand distribution of key species at lower trophic levels(Richardson & Schoeman 2004, Perry et al. 2005). Thelong-term biological consequences of acidificationare still poorly understood, but organisms reliant oncalcium carbonate exoskeletons are likely to beseverely impacted as pH decreases (Orr et al. 2005).In polar regions, the physical changes related to thegradual disappearance of sea ice are also likely tohave far-reaching ecological implications (Hunt et al.2002, Stempniewicz et al. 2007, Gaston et al. 2009).All these expected changes will mean that thefavoured prey organisms of many seabirds willchange in abundance or shift their distributions, andthis will inevitably affect seabird energy balance,demography and ultimately abundance and distribu-tion. In temperate and polar regions, such indirecteffects are likely to be more important than directphysiological effects for most species, and seabirdresponses to climate change may vary according todiet (Kitaysky & Golubova 2000).

It is often assumed, in the present state of the worldocean, that bottom-up control is the most importantmechanism structuring marine ecosystems (but seePauly & Maclean 2003). However, top-down controlis likely to also be important under certain condi-tions, and indeed it is possible that climatic changesmay affect the balance between these 2 mechanisms(Hunt et al. 2002). This implies that changing seabirdand other upper predator populations under somecircumstances can affect the abundance and/or dis-tribution of their prey, and that this can have far-reaching ecological implications, even those thatdeviate from traditional trophic cascade patterns

106

Lewison et al.: Research priorities for seabirds 107

(Cury et al 2000, Nicol et al. 2010). It is thereforeimportant that model representations of trophiceffects rely on empirical information on the relativeimportance of bottom-up and top-down control,rather than on standard assumptions.

Understanding these trophically mediated effectsand predicting their consequences requires collabo-ration with researchers working in physical and bio-logical oceanography, marine ecology and fisheriesscience. Unfortunately, in many parts of the worldthere is a lack of relevant long-term, large-scale insitu data sets, although there are notable exceptions(e.g. the Continuous Plankton Recorder survey, Reidet al. 2003). Satellite telemetry data can partly fill thisgap, as they allow estimation of, for example, pri-mary productivity with a high spatial resolution andcoverage (Behrenfeld & Falkowski 1997). Appropri-ate statistical modeling tools such as Bayesianapproaches are needed to establish the strength oftrophic links and predict consequences for seabirds(Clark 2005).

5.3. How do interactions at different hierarchicalscales affect seabirds’ responses to environmental

change?

Interactions with conspecifics in seabird coloniesare widely viewed as accruing net fitness benefits,for example through enhanced protection frompredators (Birkhead 1977). Under conditions of lowprey availability, parents may need to forage simulta-neously, leaving their offspring temporarily unat-tended (Regehr & Montevecchi 1997). Except in can-nibalistic species (Hamer et al. 1991), suchunattended chicks may be expected to gain protec-tion within colonies from the proximity of neighbour-ing birds, providing parents with a buffer againstadverse foraging conditions. However, recent evi-dence has indicated that in the absence of cannibal-ism, chicks left unattended by their parents maynonetheless be attacked and killed by adults fromneighbouring sites (Ashbrook et al. 2008) in additionto being attacked by non-breeders attempting tousurp sites (Anderson et al. 2004, Hamer et al. 2007).These data indicate potential shifts in the cost-bene-fit of breeding at high density (Danchin & Wagner1997), especially when food conditions are poor andsuggest a growing need to understand how environ-mental changes affect behavioral and social interac-tions within colonies.

Some of the most powerful data concerningresponses of seabirds to different environmental

pressures have derived from long-term studies offocal populations (e.g. Frederiksen et al. 2004, Gros-bois & Thompson 2005, Österblom et al 2007, Mon-tevecchi 2009). Few studies, however, have consid-ered responses to environmental forcing in thecontext of interactions within population networks,which may have important population dynamic con-sequences (see Coulson 1991 for an exception). Forinstance, Sandvik et al. (2008) reported a positiverelationship between sea surface temperature (SST)and the breeding success of a wide range of NorthAtlantic seabirds. For several of the species includedin this analysis, however, warming also had a nega-tive effect on annual adult survival (Grosbois &Thompson 2005, Sandvik et al. 2005, Votier et al.2005). In a circumpolar analysis of breeding murrepopulations, Irons et al. (2008) demonstrated that theArctic species, thick-billed murre Uria lomvia,responded most positively when SST warmedslightly and that the temperate species, commonmurre U. aalge, exhibited rapid increases when SSTcooled moderately. Most interestingly, both speciesresponded negatively to large shifts in SST regard-less of whether the change was in the positive or neg-ative direction.

Sites with high breeding success are likely to beattractive to prospecting non-breeders (Danchin etal. 1998), although this may not lead to reproductivesuccess after recruitment if these birds experiencerelatively low survivorship post-recruitment. Forseabirds, adult survival has much greater elasticitythan annual breeding success, which suggests thatsome breeding sites may act to some extent as popu-lation sinks (Hamer 2010), notably if predators affectthe survival of individuals at some locations. Forinstance, a recent study in Alaska, USA, found a pos-itive relationship between SST and the breeding suc-cess of piscivorous auks (Alcidae) but a simultaneousdecline in breeding population sizes (Slater & Byrd2009). It is not clear whether these declines occurreddespite increased recruitment from other colonies.This suggests that meta-population and source-sinkdynamics will become increasingly important fea-tures of seabird conservation efforts in the future(Lebreton et al. 2003, Cam et al. 2004, Zador et al.2009).

An additional challenge to understanding theeffect of climate change on seabirds is the differen-tial rates of environmental change in marine andterrestrial environments. For instance, impacts ofclimate change on prey availability at sea and directphysiological impacts on birds at the nest may showdifferent patterns of temporal and spatial variation,


leading to an increasing disparity between whereand when birds can nest and where and when theycan locate sufficient prey to feed themselves andtheir offspring (Grémillet & Charmantier 2010).Direct physiological effects have typically beenviewed as relatively unimportant (Durant et al.2004) but are likely to increase in prominence withcontinuing climatic warming in most regions, lead-ing to a growing need for combined studies of bothtypes of effect and their synergistic interactions topredict how the geographical distributions of differ-ent species are likely to alter in response to continu-ing climate change (Oswald et al. 2008). Forinstance, several statistical approaches are availableto describe the climate envelopes of different spe-cies and predict how these envelopes will shiftunder different climate change scenarios (Huntleyet al. 2007). Further work is required, however, bothto refine the models themselves, i.e. to include bothdirect and indirect effects, and to refine their pre-dictions by incorporating information on dispersal,recruitment and colony formation processes in addi-tion to other factors likely to impede or facilitategeographical range shifts (Mustin et al. 2007,Grémillet & Boulinier 2009). In marine systems withrelatively low productivity, annual stochasticity inriver flows is known to influence the populationdynamics of small pelagic fish, which are the mainprey for many seabird species (Lloret et al. 2004).Hence, in such systems the interactions between cli-mate change, human perturbations in terrestrialecosystems (e.g. presence of dams and pollutants inmajor rivers) and seabirds may be particularlystrong. Integrated research which takes into consid-eration the natural and altered patterns of spatio-temporal variability of the environment as well asthe mechanistic processes of population responses(including ecophysiology, demography, and behav-ior) is required.

6. Managing anthropogenic impacts (invasivespecies, contaminants and protected areas

As a consequence of the increasing intensity anddiversity of threats that they face both on land and atsea, seabirds have become increasingly threatenedat a faster rate than any other bird group in recentdecades (Boersma et al. 2002, Butchart et al. 2004).One of the most relevant direct threats at sea, fish-eries, was covered in Section 4. However, there areother anthropogenic impacts that also have a sub-stantial effect on seabird populations, namely inva-

sive species and pollution. Here, we also focus partic-ular attention on one important conservation toolused to address many of these threats, the implemen-tation of protected areas.

6.1. What are the population-level impacts ofinvasive species?

That seabirds have not evolved to respond to terrestrial predation by vertebrate predators is awell-known phenomenon. Most seabirds breed onislands or isolated patches to avoid terrestrial pre -dators. Unfortunately, alien fauna associated withhumans and marine traffic trade has accessed iso-lated breeding areas; this has led to drastic reduc-tions, and extinctions, in some seabird populations.Direct predation by non-native animals (mainly rats,feral cats and other carnivores) is the most studiedtype of invasive species; however, there are othersless known but equally deleterious, such as invasivegrazers or plants, which can also have negativeimpacts on breeding habitat suitability by causingchanges in vegetation structure, trampling of nests orerosion (see review in Boersma et al. 2002).

Several factors influence the level of impact of non-native species on seabirds: first, the body size of boththe predators and the seabirds. Feral cats or othermedium-sized carnivores (such as Mustelidae) canprey on a large range of seabird species (includingadults) from storm petrels to albatrosses, and thuscan have a drastic effect on their populations (Cour-champ et al. 2003). Because they are able to gainaccess to small burrows, rats and mice can have agreater impact on smaller species, e.g. terns andsmaller petrels, by preying on nests and adults, andthe presence of non-native carnivores is incompati-ble with viable populations of these seabirds (Martinet al. 2000, Angel et al. 2009). Rats and mice (Cuth-bert & Hilton 2004, Wanless et al. 2009) can also preyon eggs and small chicks of larger seabirds — evenalbatrosses — although many of these populationsare able to buffer this impact. The best example ofseabird resilience to an invasive predator is thecohabitation of shearwaters and rats on Mediter-ranean islands (Ruffino et al. 2009). The world’slargest colony of Leach’s storm petrel Oceanodromaleucorhoa on Baccalieu Island in the northwestAtlantic has been co-inhabited by red foxes Vulpesvulpes for about a century (Montevecchi & Tuck1987). The foxes prevent large ground-nesting gulls,which also take a large toll of storm petrels (Sten-house et al. 2000), from nesting on the island.

108


Knowing which life-history stage is most affectedby the presence of an invasive species is also crucialto determining the rate and severity of the impact onseabird population dynamics: invasive plants, graz-ers (such as rabbits, or cattle) or rodents mainly affectreproduction (through reduced nest site availabilityand/or quality or decreased breeding success), andtheir impact tends to be relatively low compared tothat of alien species which reduce adult survival (seeSection 1.1). Some species can buffer such distur-bances by dispersing to safer sites, particularly thosespecies with nomadic strategies, such as some gullsand terns (Oro et al. 1999), whereas less plastic spe-cies may be less able to react to predators (Igual et al.2007, but see Bonnaud et al. 2009). Finally, thebreeding habitats of seabirds can also influence theimpact of non-native species: burrowing species,compared to open-habitat species, are better pro-tected against mice, rats and relatively large carni-vores (Igual et al. 2006, Bourgeois & Vidal 2007).Cliff-nesting species are the best protected againstall types of invasive species owing to their relativelarge and inaccessible nests.

Seabirds are similarly threatened by invasive plantspecies, which have been documented to negativelyimpact breeding activity (Van der Wal et al. 2008). In-vasive plants reduce nesting success by decreasingavailable nesting habitat, entrapping or obscuringchicks, and by increasing the abundance of insectsthat may threaten eggs or chicks. In addition to pre-venting breeding or curtailing the success of breedingbirds, by reducing seabird abundance in general bothplant and mammal invasive species can alter nutrientcontent and soil composition, causing widespread ef-fects across an entire area (Hawke & Clark 2010).

The best means of controlling alien species is toreduce their numbers, or when possible, to eradicatethem (Courchamp et al. 2003). To assess the successof these programs, studies evaluate correlatedchanges in seabird parameters such as breeding out-put and adult survival, with different densities ofalien species. This can be based on direct observa-tion (Nogales et al. 2004, Igual et al. 2006) or can usenon-invaded sites as controls (Ratcliffe et al. 2010).Because of the complex community interactionsinvolved, eradications of invasive species may haveunexpected results, e.g. the substitution of an eradi-cated predator by an ecological competitor occupy-ing the same niche (Courchamp et al. 2003,Martínez-Abraín et al. 2004, Rayner et al. 2007).Nonetheless, eradication and invasive species con-trol remains an important conservation issue forseabird colonies worldwide.

6.2. What is the population-level influence of conta-minants and other pollutants?

Human activities generate a range of pollutantsthat are widely distributed across the seas, becauseof the unfettered nature of transport in aquatic envi-ronments (see Halpern et al. 2008). Pollutants includehydrocarbons, heavy metals, hydrophobic persistentorganic pollutants (POPs), and small plastic debris.Since seabirds are at or near the top of marine foodwebs, they are particularly sensitive to these pollu-tants (see review in Burger & Gochfeld 2002). For thesame reason, seabirds are excellent tools to monitortemporal and spatial trends of pollutants in themarine environment (Furness 1993).

Oil spills are a prevalent common cause of anthro-pogenically caused mass mortality in seabirds.Munilla et al. (2007) reported as many as 20 ship-wrecks involving the release of oil or other hazardousproducts into the Atlantic waters of the IberianPeninsula from 1957 to 2002. Since 1948, at least 13major petroleum spills have caused direct mortalityin at least 500 African penguins (Underhill 2007).Some of these catastrophes have had dramaticeffects on breeding seabird populations due to directmortality and depletion of food sources (Velando etal. 2005a,b, Votier et al. 2005, 2008b, Martínez-Abraín et al. 2006).

Pollutants affect seabirds in many different ways,at both individual and population levels (Burger &Gochfeld 2002). Direct mortality is the most obviouseffect, particularly when it is related to point-sourcepollution such as oil spills and illegal discharges ofoily bilge water and tank flushings, which can killlarge numbers of birds in a short time period (Piatt etal. 1990). However, sub-lethal effects can also bevery important, affecting development, physiologyand behavior, and ultimately reproductive perfor-mance and survival rates (Finkelstein et al. 2006).Pollutants can also affect seabirds indirectly by alter-ing their habitat structure and prey availability.Assessing sub-lethal effects on an individual requiresa combination of laboratory and field studies. Exam-ining the toxic residues, their metabolites and bio-chemical markers indicating the stress and tissuedamage is a useful tool (Burger & Gochfeld 2002).

Assessing the impact of pollutants on seabirds atthe population level is also extremely challenging.The population decreases seen in some seabird spe-cies are believed to be associated with exposure tohigh levels of dichlorodiphenyl dichloroethylene(DDEs) and polychlorinated biphenyls (PCBs) inNorth America during 1950s and 1960s (Hatch &

109


Kubiak et al 1989, Weseloh 1999, Nisbet 2002),although direct tests are lacking. Even in the case ofacute pollution events causing large numbers ofcasualties it is difficult to directly link a pollutant tomortality because of compensatory changes inrecruitment, reproductive and survival rates (Wiens1996) and changes in prey availability caused bysuch acute events (Peterson et al. 2003, Velando et al.2005a). The analysis of long-term population datasets can shed some light on this topic (Votier et al.2005, 2008b, Martínez-Abraín et al. 2006), especiallythose studies that have focused on vulnerability todifferent levels of pollution of POPs and plastics(Finkelstein et al. 2006, Young et al. 2010).

6.3. How effective are protected areas in protectingseabirds?

Historically, seabirds have received protection inmany of their breeding colonies across the world, andhave benefited greatly from it (Grémillet & Boulinier2009). Indeed, the designation of seabird nestingsites as protected areas has prevented direct humanimpacts such as harvest and habitat destruction.Moreover, active management focused on some ofthese places has also helped to reduce threats posedby non-native and invasive species, human visita-tion, and disease introduction (Weimerskirch 2004,Rauzon 2007). This has resulted in positive effects atthe population level (Furness & Monaghan 1987,Grandgeorge et al. 2008), although current seabirddeclines worldwide related to threats at sea demon-strate that complementary conservation strategies,including the implementation of Marine ProtectedAreas (MPAs), are necessary (Hyrenbach et al. 2000,Grémillet & Boulinier 2009).

The protection of marine areas for seabirds haslagged behind terrestrial sites, with efforts so far con-centrated in the areas adjacent to seabird colonies(Airamé et al. 2003, Wilson et al. 2008, Yorio 2009,Pichegru et al. 2010) and, to a lesser extent, in coastaland/or shallow areas for highly aggregated winter-ing species (Skov et al. 2007). However, the rapiddevelopment of tracking technologies, along with theproliferation of boat-based seabird surveys and thedevelopment of analytical tools to assess distributionpatterns and habitat preferences, has recentlyattracted attention to MPA design for seabirds andother marine megafauna, particularly in pelagicwater (see Section 2, Hooker & Gerber 2004, Hyren-bach et al. 2006, Louzao et al. 2006, 2009, Skov et al.2008, Game et al. 2009). Ongoing research has been

aided by the international commitment to protect atleast 10% of each representative habitat of the globeby 2010/2012, including the high seas, under theCBD (Coad et al. 2009). Within this framework,BirdLife International has chosen as a priority theextension of its IBA Programme to the marine envi-ronment (BirdLife International 2004), providing anexcellent opportunity for the inventory of seabirdhotspots at a global scale. Pioneering examples arethe recently completed inventories of marine IBAs inPo

research priorities for seabirds: improving conservation and … · 2. at-sea spatial ecology 2.1....

Documents