Sensory-based conservation of seabirds: a review of management strategies and animal behaviours that facilitate success

I. INTRODUCTION

Conservation biology traditionally focusses on reducing key threats such as habitat loss or overfishing (Soulé, 1985). More recently, conservation has begun to integrate evolutionary biology and the natural behaviours of target species (Crandall et al., 2000). Sensory ecology offers a powerful opportunity to influence animal behaviour for management purposes because acoustic, olfactory, and visual cues and signals are fundamental in many interactions with conspecifics, prey, predators, the environment, and human activities. However, sensory-based conservation efforts often produce variable results and can lack adequate controls. This popular and emerging sub-discipline of conservation biology merits rigorous review to improve success rates, and importantly, to facilitate application of effective techniques to a wider range of at-risk species by identifying key behavioural and life-history traits associated with successful conservation attempts.

Seabirds are excellent candidates for sensory-based conservation because they employ a range of sensory modes, display complex behaviours and cognition (Schreiber & Burger, 2001), and are often highly threatened (Croxall et al., 2012; IUCN, 2014; Paleczny et al., 2015). Seabirds have experienced steep recent population declines – 28% of the ∼347 seabird species are globally threatened and 5% are critically endangered (Croxall et al., 2012). Key orders include the Procellariiformes (tube-nosed seabirds), Sphenisciformes (penguins), Suliformes (gannets and boobies), some species of Charadriiformes (e.g. terns), Pelecaniformes (e.g. pelicans), and Phaethontiformes (tropicbirds). Of these orders, the Procellariiformes and Sphenisciformes are the most threatened, both on land and at sea (Croxall et al., 2012). On land, colonies succumb to residential and recreational development (Beatley, 1991), and even protected sites are nearly always threatened by physical disturbance or invasive predators (rats, cats, dogs, and mustelids; Burger & Gochfeld, 1994; Micol & Jouventin, 1995). At sea, one of the greatest threats is incidental bycatch by fishing vessels, especially when ensnared by hooked baits and longlines (Brothers, 1991; Murray et al., 1993; Tasker et al., 2000; Taylor, 2000; Bull, 2007; Richard, Abraham & Filippi, 2011). Estimates exceed 160000 seabirds killed per year across 68 fisheries (Anderson et al., 2011). These threats make seabird conservation a priority, ideally via practical, cost-effective techniques that can be applied generally across many species. Fortunately, seabirds' shared life-history traits and sophisticated sensory systems mean that sensory-based conservation techniques are a strategic option.

Traditionally, seabird conservation managers have focused on pest eradication, chick translocations, fisheries bycatch reductions, and social attraction to restored nesting sites (e.g. Parker et al., 2007; Jones & Kress, 2012). Although rarely explicitly recognized, sensory-based techniques are often used (Table 1) with variable success (see Section 4). Attracting birds to secure nesting habitat often involves broadcasting vocalizations or installing decoys to suggest a burgeoning colony (Kress, 1998; Gummer, 2003; Miskelly & Taylor, 2004; Table 1). Bycatch reduction techniques involve equipping fishing vessels with unpleasant sounds or vibrant visual banners, as well as changing the timing of line setting and hook weighting (Bull, 2007). Again, effectiveness varies (Table 1), especially based on species' responses to specific threats (for a review see Melvin et al., 1999). Despite the popularity of these sensory-based conservation efforts, there is little recognition that their success depends upon evolved behaviours and sensory ecology, and that measuring success requires comparison to unmanipulated controls.

Seabirds have evolved a variety of unique physiological and sensory adaptions in response to the selective pressures imposed by their pelagic and colonial lifestyle. Seabirds live almost entirely at sea where foraging and navigation requires long-distance sensory modes that are less apparent in non-pelagic birds (Stager, 1967; Hunt & Schneider, 1987; Boyd & Croxall, 1996; Jarvis et al., 2014). For example, many seabird taxa can see fish under the ocean surface (e.g. albatross; Martin, 1998) or forage via chemical cues (e.g. Procellariiformes and Sphenisciformes; Nevitt, Veit & Kareiva, 1995; Cunningham, Strauss & Ryan, 2008; Pollonara et al., 2015). Natal philopatry and colonial breeding mean many seabird species must also navigate to, assess and recognize their breeding colonies, which are often on isolated islands (Schreiber & Burger, 2001). Some species use audition; playback experiments with recorded bird calls can emulate this (e.g. for Galápagos petrels Pterodroma phaeopygia; Podolsky & Kress, 1992). Seabirds may also use scents to locate colonies (Grubb, 1973; Bonadonna, Spaggiari & Weimerskirch, 2001). For example, when Cory's shearwaters (Calonectris borealis) were made anosmic, orientation and navigation were impaired and they could not return to their colony when displaced (Gagliardo et al., 2013).

Seabird colonies may have thousands of birds (Schreiber & Burger, 2001), selecting for mechanisms to identify mates or nests within a crowd. Some seabirds can use vocalizations to find chicks after foraging bouts (e.g. Adélie and gentoo penguins, Pygoscelis adeliae and P. papua; Aubin & Jouventin, 2002). Visual discrimination is unlikely for nocturnal seabirds such as Procellariformes, which instead can use olfactory signals to identify species or individuals (Bonadonna & Nevitt, 2004; Bonadonna et al., 2007; Bonadonna & Mardon, 2010; Mardon et al., 2010). A classic test with captive nocturnal Leach's storm petrels (Oceanodroma leucorrhoa) found they were more likely to orient to the smell of conspecific nesting material than a control forest odour (Grubb, 1974), although the role of individual versus conspecific odours in this experiment is unclear. Chemosensory communication is not necessarily restricted to nocturnal species; diurnal black-legged kittiwakes (Rissa tridactyla; Laridae) have both individual and sex differences in odour (Leclaire et al., 2011).

Another seabird life-history trait selecting for sophisticated sensory behaviour is monogamy and the formation of long-term pair bonds, necessitating accurate assessment of conspecific quality during mate choice (Warham, 1990; Schreiber & Burger, 2001). Breeding birds often use vocal signals to differentiate between sexes (Brooke, 1978; James, 1984; Storey, 1984; Taoka & Okumura, 1990). Amongst closely related shearwater species, both sex- and species-specific vocalizations are common (Curé et al., 2012). For example, the exhalent, low part of the call is sufficient for species discrimination between Yelkouan (Puffinus yelkouan), Balearic (P. mauretanicus) and Cory's shearwaters, and is likely also important for sexual differentiation (Curé et al., 2012). By contrast, in the Australasian gannet (Morus serrator), male and female vocalizations do not appear to differ (Krull et al., 2012), but plumage colour does when modelled into a bird visual system (Ismar et al., 2014). Mate choice via assessment of visual signals is well documented, e.g. the brightly coloured legs of the blue-footed booby (Sula nebouxii) indicate individual quality, and affect mate choice, and therefore male breeding success (Velando, Beamonte-Barrientos & Torres, 2006). Similarly, male frigatebirds (Fregata spp.) use distinctive red gular pouches in sexual displays (Juola, McGraw & Dearborn, 2008). Evidence for mate choice via odour cues is rare for seabirds, however the major histocompatibility complex (MHC), which is associated with immune function and individual odours, influenced mate choice in the blue petrel (Halobaena caerulea; Strandh et al., 2012). In short, seabirds have evolved unique behaviours which in turn require sophisticated sensory mechanisms. These involve several signal modes, conspecific and heterospecific signals, and long- and short-distance signal reception – all of which could be harnessed for conservation and management.

Although we focus on seabirds, our findings can be broadly applied to other animals with similar foraging, navigation, colonial, mate choice, and individual and breeding-site recognition behaviours. We analyse how sensory traits associated with these behaviours can be co-opted for conservation purposes, highlighting which approaches and taxa yield the greatest conservation benefits. We begin by reviewing the seabird literature to describe and quantify the range of sensory-based conservation techniques, factors associated with success, and the general potential for sensory-based conservation biology. We assess research effort according to seabird taxonomic groups (Section 2, Fig. 1), and the sensory mode manipulated (auditory, olfactory, visual, or combined; Fig. 2). Importantly, we provide a summary list of behaviours typically associated with successful sensory-based conservation efforts in seabirds, which can be broadly applied to other at-risk taxa. An understanding of the uses and ecological basis of sensory-based techniques is critical for improving conservation efforts, and heralds a new, more evolutionary approach to species management.

II. METHODS

We searched for publications on seabird conservation and management efforts that explicitly involved the target species' auditory, visual, or olfactory senses using two databases (Google Scholar, Scopus) in English, with no limit on year. Database searches were made using a combination of the words: auditory, vocalization, playback, visual, decoy, olfactory, odour, sensory AND seabird conservation (N studies yielded in Google Scholar = 18500 of which first 3000 were searched based on relevance, N studies yielded in Scopus = 158; total N of studies included from both databases = 13 plus 2 reviews) OR seabird colony attraction (N studies yielded in Google Scholar = 7230 of which first 2000 were searched, N studies yielded in Scopus = 61; total N studies included from both databases not included in previous search = 4) OR seabird bycatch (N studies yielded in Google Scholar = 3890 of which first 1000 were searched, N studies yielded in Scopus = 26; total N studies included from both databases not included in previous search = 2) OR seabird translocation (N studies yielded in Google Scholar = 1820, N studies yielded in Scopus = 55; total N studies included from both databases not included in previous search = 4) OR seabird bycatch (N studies yielded in Google Scholar = 3890, N studies yielded in Scopus = 26; total N studies included from both databases not included in previous search = 2). Articles selected were read based on the title and abstract from the search, and then included in this review based on the content of the article. We excluded studies focused on censusing populations, sensory methods for deterring pest seabirds (e.g. gulls at landfill sites), or the sensory behaviour of other species interacting with seabirds (e.g. deterrents for invasive predators that eat birds). Further studies were included based on work reviewed by Jones & Kress (2012) and Melvin et al. (1999). We systematically searched the databases and aimed to incorporate all studies relevant for our study, but may have inadvertently excluded papers due to our specific search parameters.

We then quantified the number of sensory-based conservation publications per seabird order (Fig. 1), and the sensory modes utilized per study (auditory, olfactory, visual, or multi-modal; Fig. 2). Based on our literature review (summarized in Table 1), we performed a logistic regression with chi-squared analysis of deviance comparing the sensory mode targeted in conservation efforts (acoustic, visual, olfactory, or multi-modal) based on the seabird order of targeted species (Procellariiformes or Charadriiformes; Suliformes were not included as only one study was reported). We hypothesized that species behaviours (reproductive habitat and activity times) influence the success of sensory-based colony attraction techniques. We performed one-tailed Fisher's Exact tests on the null hypothesis that success of sensory-based colony attraction was no different based on reproductive habitat (protected habitat – cliffs and burrows, and exposed habitat – ground), and activity period on colonies (nocturnal, diurnal).

III. SENSORY MODES, TAXONOMY, BEHAVIOUR AND SUCCESSFUL CONSERVATION OUTCOMES

Sensory-based conservation techniques have not been trialled equally across seabird orders or families. Procellariiformes are the most studied (N = 14 species out of ∼125 species in the order) with most of these focusing on the family Procellariidae (N = 11). Charadriiformes are the second most common order represented in the literature for sensory-based conservation (N = 10 out of ∼133 species). There have been no sensory conservation studies on Phaethontiformes or Sphenisciformes, and only one study on a suliform species (Fig. 1). Sensory techniques could and should be attempted across a broader taxonomic range, but this is impeded by the lack of data about sensory ecology in the neglected groups. The Sphenisciformes in particular should be a high priority for sensory research and sensory-based conservation because species in this order are some of the most critically threatened seabirds (Croxall et al., 2012).

Although there was no correlation between the sensory mode targeted and the likelihood of a successful conservation outcome, taxonomy did correlate with success (Table 2). Sensory-based conservation efforts are more likely to succeed with Procellariiformes than Charadriiformes (Table 2; Fig. 2). This is likely due to differences in the typical activity patterns and nesting sites for these two orders – Procellariiformes are typically nocturnal and burrow-nesting whereas Charadriiformes are generally surface-nesting. We found some possible trends between nocturnal activity time and protected (burrow and cliffs) versus exposed (surface/ground) breeding, but results were not significant (Fig. 3). These results are likely biased by the under-reporting of failed conservation attempts (Fanelli, 2011).

IV. USE AND EFFECTIVENESS OF SENSORY MODES IN CONSERVATION

V. BEHAVIOURS FACILITATING SENSORY-BASED CONSERVATION

Since sensory approaches to conservation rely on natural interactions between target animals and their conspecifics, heterospecifics and the environment rather than any specific individual species traits, lessons from seabird efforts can be applied to a broad range of at-risk animal taxa. Sensory-based conservation takes advantage of species' natural behaviours. Animals clearly differ in their sensory ecologies and amenability to sensory manipulation for management outcomes. However, our survey has revealed several life-history traits that characterize a species as particularly suitable for sensory-based conservation.

VI. COGNITIVE CHALLENGES FOR SENSORY-BASED CONSERVATION

A nuanced sensory-based approach to conservation must consider ways in which an animals' cognition may limit sensory management efforts. The tendency of animals to habituate to sensory treatments is especially important because deterrents, unlike true threats, do not impose negative consequences on the receiver. Playing different combinations or cycles of stimuli might prolong deterrent effects and minimize habituation, as well as decrease interference with other non-target species. In tests with harbor porpoises (Phocoena phocoena), some habituation to sounds occurs with sound cycling, but intermittent pings were still relatively successful (Carlström, Berggren & Tregenza, 2009). Use of high-frequency noise to deter tammar wallabies (Macropus eugenii) from areas of human conflict were shown to have no effect, regardless of the presence of food. This might be because the sounds were artificial and not part of the animals' natural behaviour (Muirhead et al., 2006; also shown in kangaroos Macropus spp.; Bender, 2003). This indicates the importance of investigating the natural sensory ecology and responses of target conservation species to different stimuli.

Anthropogenic sensory traps are human activities that lure animals from natural sensory cues and are associated with urbanization (e.g. artificial light; Schlaepfer, Runge & Sherman, 2002). Anthropogenic sensory traps can cause significant individual- and population-level effects, e.g. artificial light and fires disorient newly hatched sea turtles (Madliger, 2012). A well-studied example is the consumption of plastic particles in the ocean by albatrosses and other seabirds (Harper, 1979; Robards, Piatt & Wohl, 1995; Auman et al., 1997; Schlaepfer et al., 2002), effectively reducing food consumption (Ryan, 1988). A second major topic for recent research is how urban sounds interfere with bird songs, negatively impacting communication in many species worldwide (Slabbekoorn & Ripmeester, 2008). Determining when an animal is subject to ecological traps and applying appropriate mitigation measures is becoming a more widely recognized aspect of sensory-based conservation applicable across various taxa (Madliger, 2012). For example, anthropogenic noises, such as traffic sounds that interfere with predators' ability to detect prey could be modified by reducing the speed limit (Madliger, 2012). By considering the sensory perceptions of target species, conservation biologists can predict where such traps may exist and use their knowledge of the animals' sensory modes to alleviate the impact of traps.

Another risk of sensory-based conservation efforts is negative effects on unintended receivers or eavesdroppers. For example, acoustic pingers to deter seals from aquaculture pens also deter non-target harbor porpoises (Johnston, 2002). Another risk of eavesdroppers involves predators and parasites, which find prey via multiple sensory modes, including chemical (Stowe et al., 1995) and auditory cues (Deecke, Ford & Slater, 2005). The risk of attracting predators and parasites may be intensified if management efforts broadcast sensory stimuli, such as playing vocalizations for translocation. While the effect of eavesdroppers has not been thoroughly examined in sensory-based conservation examples, we recommend that predator and parasite activity be carefully monitored before and during sensory-based conservation activities.

VII. CONCLUSIONS

(1) Sensory-based conservation incorporates behavioural and sensory ecology by considering how signalling and responses evolve under selection from conspecifics, heterospecifics and the environment. Conservation approaches informed by evolutionary principles are likely to have several advantages over more-interventionist methods such as captive breeding and catch limits as they are transferable among taxa, and can be applied simultaneously to several species with similar behaviours (e.g. bycatch species).

(2) Sensory-based approaches may prove to be more cost-effective than interventionist approaches, especially if they can be automated (e.g. acoustic anchoring versus manual translocations).

(3) Sensory conservation methods have been successfully applied in a number of seabird taxa (e.g. Sternidae, Alcidae, Procellariidae; Table 1). While there are probably examples of sensory-based conservation in seabirds not encompassed by our search parameters, we found examples of attraction attempts that had limited success (e.g. Audouin's gulls, Ichthyaetus audouinii; Table 1), and other groups whose responsiveness to such techniques remains seemingly unstudied (e.g. booby and gannet species, frigatebirds, and tropicbirds; Jones & Kress, 2012; Fig. 1).

(4) This review identifies several conservation contexts where sensory-based techniques have considerable potential to improve outcomes such as establishing new colonies and deterrence from anthropogenic traps. We recommend:

(a) Life-history and behavioural traits that select for sensory signalling to inform the best candidate species for sensory-based conservation. These traits are: long-distance travel, navigation in featureless environments (e.g. seas, deserts), pair-bonding and mate choice, high parental care by both sexes, conspecific and heterospecific recognition (prey and predators, and mates, nests and chicks), coloniality, philopatry.

(b) Using multiple sensory modes to promote stronger behavioural responses from the target species. For example, combining sound and visual lures could be more effective than either lure on its own when attempting to counter effects of strong philopatry.

(c) Evaluating the potential risks of unintended consequences from using sensory-based methods. Predator cues should be investigated to avoid luring seabirds, or other prey species, to ecological traps, or attracting predators to the target species.

(d) Incorporating an experimental control to improve the ability to evaluate conservation outcomes and quantifying success of management efforts. For example, studies on attracting seabirds to new locations rarely observe other suitable nearby habitat without experimental stimuli to ensure that the movement of birds to the new location is a result of sensory method and not just a coincidental outcome of animal movement.

(e) Undertaking additional research on sensory-based approaches to conservation of seabirds and other taxa to optimize conservation outcomes. In seabirds, olfaction is underutilized. Further work (and controlled experimentation) on visual attractants is required to understand why some attempts are successful and others are not. A high priority is to understand the ethological response of seabirds to different vocalizations, such as identifying which calls are most effective in enticing seabirds to colony locations.

(5) Systematic investigations into sensory ecology could profoundly improve the success of conservation strategies. This type of sensory management is broadly applicable to many systems, and may be more cost-effective than more interventionist activities.