Visual Fields in Blue Ducks Hymenolaimus

Ibis (2007), 149, 112–120

Blackwell Publishing Ltd Visual ﬁelds in Blue Ducks Hymenolaimus malacorhynchos and Pink-eared Ducks Malacorhynchus membranaceus: visual and tactile foraging

GRAHAM R. MARTIN1*, NIGEL JARRETT2 & MURRAY WILLIAMS3† 1Centre for Ornithology, School of Biosciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK 2Wildfowl and Wetlands Trust, Slimbridge, Gloucestershire GL2 7BT, UK 3Department of Conservation, PO Box 10420, Wellington, New Zealand

Blue Ducks Hymenolaimus malacorhynchos (Anatidae), an IUCN Red Listed Endangered species, reside in headwaters of New Zealand rivers and feed primarily on aquatic invertebrates. However, whether such food items are detected by tactile or visual cues is unknown. That Blue Ducks may use tactile cues when foraging is suggested by the presence of specialized flaps of thickened, keratinized epidermis containing Herbst’s corpuscles along the ventral margins of the upper mandibles near the bill tip. Similar bill flaps are found only in one other duck species, Pink-eared Ducks Malacorhynchus membranaceus, that surface filter-feed on a range of planktonic organisms. Using an ophthalmoscopic reflex technique we determined the visual fields of both species. In Blue Ducks the eyes are frontally placed resulting in a relatively wide binocular field into which the narrow tapering bill intrudes. There is a large blind area to the rear of the head. This visual field topography is similar to that of other visually guided foragers including those that take mobile prey from the water column, e.g. penguins (Spheniscidae). By contrast, Pink-eared Duck visual fields show features found in other tactile feeding ducks: a narrow frontal binocular field with the bill falling at the periphery, and comprehensive visual coverage of the celestial hemisphere. We conclude that although Blue Ducks may take prey from rock surfaces they are primarily visual feeders of the water column and we suggest therefore that their foraging may be significantly disrupted by changes in water clarity. This introduces a previously unconsidered factor into the selection of sites for population enhancement or re-introductions, a current conservation focus.

In birds, the position and extent of the region of visual cues is thought to be primarily a result of the binocular vision appear to be determined primarily symmetrical optical flow-fields generated in each by feeding ecology (Martin & Katzir 1999). Of prime eye within the forward-facing binocular sector, as the importance is the degree to which vision is used for head moves towards an object (Martin et al. 2005). the precise control of bill position when pecking or Such flow fields specify directly the direction of lunging at prey, or when feeding chicks. In species travel and the time to contact a target, and this infor- that feed in this way the bill falls either centrally or mation is processed sufficiently fast by the brain to just below the centre of the frontal binocular field control rapid pecking or lunging at items. Stereo- (Martin et al. 2004, 2005). The interpretation of this scopic cues that may be available from binocular arrangement is that the control of bill position by vision are, however, too slow for the control of such rapid movements (Davies & Green 1994). *Corresponding author. In birds that do not require such precise control of Email: [email protected] rapid bill movements (probe and filter feeders with †Present address: School of Biological Sciences, Victoria University precocial self-feeding chicks), the bill falls outside of Wellington, PO Box 600, Wellington, New Zealand or at the periphery of the binocular field. In such

Visual ﬁelds and foraging in ducks 113

examples the binocular field extends above and behind point of attachment to the upper jaw (Kear & Burton the head, and so provides comprehensive visual 1971). It is this structure that is referred to in the coverage, but it is clear that vision cannot be used specific name. Among birds a similar structure is found for the precise control of bill position (Martin 1994, only in one other duck species, Pink-eared Ducks Guillemain et al. 2002). that feed exclusively by filter-feeding to obtain a range Knowledge of visual field characteristics in birds of planktonic organisms (algae, microscopic seeds, can therefore be used to determine the importance crustaceans, molluscs, insects) in shallow, still and of visual and tactile cues in their foraging. We have typically turbid waters (Kear 2005). The structure of employed this technique to determine whether Blue these flaps has not been fully described but they Ducks Hymenolaimus malacorhynchos (Anatidae) appear to be well endowed with Herbst’s corpuscles use visual and/or tactile cues to guide their foraging. and are thought to have a role in the tactile feeding For comparison we also determined visual fields in (Kear & Burton 1971). However, bill size and shape Pink-eared Ducks Malacorhynchus membranaceus, a differ markedly in these two duck species (Fig. 1). species whose foraging is known to be primarily The bill of Pink-eared Ducks is typical of many guided by tactile cues (Kear 2005). Anatidae in that it is expanded anteriorly. However, New Zealand’s Blue Ducks presently inhabit forested the bill of Blue Ducks is quite different, being narrow headwater catchments of rivers with medium to steep and tapering towards the tip, and with a straight gradients (Kear & Burton 1971, Collier et al. 1993). culmen; these are a group of structural adaptations Stomach analysis has shown that they feed on a wide associated in other bird taxa with obtaining food variety of aquatic invertebrates (Collier 1991, Veltman items by pecking or lunging at individual items (Kear et al. 1995) and occasionally on small amounts of & Burton 1971). plant material (Marchant & Higgins 1990). How Here we show marked differences between the these birds catch their prey is not known. Are these visual fields of Blue Ducks and Pink-eared Ducks. food items detected by tactile cues when ‘blind Pink-eared Duck visual fields show clear charac- searching’ among rocks and stones of the riverbed, teristics of a tactile feeder whereas Blue Ducks have or are Blue Ducks detecting individual items by sight visual fields which suggest that they are primarily either on the substrates or in the water column? This visually guided foragers. is an important question; Blue Ducks are the subject of considerable conservation concern because of their METHODS ongoing population decline and range contraction. They are classified as an IUCN Red Listed Endangered We determined visual fields in two adult Blue Ducks species with an estimated world population of 2440 and two adult Pink-eared Ducks (sexes unknown) individuals (BirdLife International 2006). Population from the captive breeding collections held by the decline and range contraction were attributed initially Wildfowl and Wetland Trust Centres at Slimbridge to competition for aquatic insects with introduced (Gloucestershire) and Arundel (West Sussex), England. fish, but more recent attention has focused upon the Visual field parameters were determined using an deforestation of river catchments causing a decrease ophthalmoscopic reflex technique, which has been in water clarity and silting leading to a reduction used in a range of birds of different phylogeny, ecology of invertebrate prey, and to introduced mammalian and feeding techniques and which readily permits predators (Kear 2005). If Blue Ducks are primarily interspecific comparisons (Martin & Coetzee 2004, visual feeders their foraging may be significantly Martin et al. 2005). For a description of the apparatus disrupted by changes to water clarity. This would and methods see the Appendix. The procedures used introduce a previously unconsidered factor into the were performed under guidelines established by the selection of sites for population enhancement or United Kingdom, Animals (Scientific Procedures) re-introductions, a current conservation focus (Collier Act, 1986. 2004). That Blue Ducks may use tactile cues when foraging RESULTS is suggested by the presence of the flaps at their bill tips (Fig. 1). These flaps are relatively simple struc- The positions of the visual field margins in each of tures with a heavily keratinized epidermis containing the birds were within 5° of each other at all elevations a small number of Herbst’s corpuscles (Gottschaldt and hence we present mean data for each species. 1985), mainly confined to the sections nearest the Maps of the frontal binocular fields are shown in

114 G. R. Martin, N. Jarrett & M. Williams

Figure 1. Heads of Blue Ducks and Pink-eared Ducks. (a,b) Frontal and lateral views of a head of a Blue Duck. (c–e) Frontal, lateral and rear views of the head of a Pink-eared Duck. In a and c the birds are photographed approximately along the plane containing the projection of the line of the bill, as shown in Figures 2 and 3. In e the bird is photographed from behind in the direction of the plane shown in Figure 3. The corneas are clearly visible. The ﬂaps of skin attached to the tip of the upper mandibles in the Blue Ducks are the heavily pigmented areas at the bill tip in a and b. Scale bar, 20 mm.

Figure 2. Horizontal sections through the visual The frontal binocular fields fields in an approximately horizontal plane are shown in Figure 3, and the width of the binocular field The maximum width of the frontal binocular field as a function of elevation in the median sagittal plane in both species occurs approximately 20° above the in the Pink-eared Duck is shown in Figure 4. These bill. However, maximum binocularity in Blue Ducks results show that the visual fields of these two (34°) is twice that in Pink-eared Ducks (17°). More- species of ducks differ markedly. These differences over in Pink-eared Ducks the bill is positioned at arise because the eyes of Blue Ducks are more the very periphery of the visual field whereas in Blue frontally placed in the skull whereas those of Pink- Ducks the bill intrudes into the binocular field and eared Ducks are more lateral and higher in the skull determines its limits, resulting in the ‘notch’ in the (Fig. 1). In addition the visual field of each eye is binocular field at the elevation of the bill (Fig. 2). narrower in Blue Ducks (173°) than in Pink-eared This indicates that a Blue Duck can see its own bill Ducks (190°) (Fig. 3). tip while a Pink-eared Duck cannot.

Visual ﬁelds and foraging in ducks 115

Figure 2. Visual fields of Blue Ducks and Pink-eared Ducks. Each diagram shows a perspective view of an orthographic projection of the boundaries of the retinal fields of the two eyes. It should be imagined that the bird’s head is positioned at the centre of a transparent sphere with the bill tip projecting towards the point shown and the field projected onto the surface of the sphere (grid at 20° intervals). The drawing adjacent to each diagram shows a side view of the head with the bill at the correct orientation.

coverage of the hemisphere about its head, but a Limit of visual fields to the rear of Blue Duck could only achieve similar visual coverage the head by the use of scanning head movements. To the rear of the head the visual fields of these two species also differ markedly. In Blue Ducks there is DISCUSSION an extensive blind area behind the head (48° wide at the approximately horizontal plane) whereas in Pink- Visual fields and foraging eared Ducks the visual fields of each eye overlap or coincide at all elevations behind the head (Figs 3 & 4). As in other tactile and filter-feeding ducks (Mallards Thus, a Pink-eared Duck has comprehensive visual Anas platyrhynchos and Northern Shovelers A. clypeata)

116 G. R. Martin, N. Jarrett & M. Williams

Figure 3. Horizontal sections through the visual fields of Blue Ducks and Pink-eared Ducks in a plane containing the frontal binocular field at its maximum width. This plane is indicated by the line through the eye in each of the drawings of the birds to the right of each visual field diagram.

(Martin 1986, Guillemain et al. 2002), Pink-eared (Martin & Coetzee 2004) that are associated with Ducks have Type 2 visual fields (Martin & Coetzee visually guided pecking or lunging at prey: (1) the 2004), which give them comprehensive vision of the bill tip projection falls either centrally or within the hemisphere that surrounds their heads (Fig. 4). As lower half of the binocular area, (2) the binocular also found in Mallards and Shovelers, maximum field is relatively long and narrow with maximum binocular overlap (17°) occurs in the frontal field binocularity of approximately 30°, (3) maximum close to the horizontal, with the bill tip falling at the binocularity occurs 20–40° above the projection of very lower periphery of the visual field. the bill tip and (4) there is a large blind area to the The visual fields of Blue Ducks, however, exhibit rear of the head. Furthermore, in Blue Ducks the bill the four principal characteristics of Type 1 visual fields intrudes into the binocular field implying that a bird

Visual ﬁelds and foraging in ducks 117

Figure 4. Binocular ﬁeld width in Pink-eared Ducks as a function of elevation in the median sagittal plane. The orientation of the bird’s head is shown diagrammatically.

can see its own bill tip and thus see items between associated with stereopsis (the perception of rela- the mandibles when they are opened. These features tive depth at close quarters that results from the are found in a wide range of species which differ in simultaneous combination of information from the their ecology and phylogeny but have in common two eyes) but is the result of having part of each eye’s precision pecking or lunging at prey. For example, monocular field extending across the sagittal plane to similar visual field topographies are found in ground- ensure that a section of each eye’s visual field looks hornbills and hornbills (Bucorvidae and Bucerotidae), in the direction of travel (Martin & Katzir 1999). which take individual items in the tips of their This arrangement provides a radially symmetrical elongated down-curved bill by precision pecking portion of the optic flow-field in each eye. From (Martin & Coetzee 2004), and in penguins (Sphenis- this, information can be extracted that accurately cidae), which take aquatic prey directly in their bill specifies both direction of travel and time to contact in rapid pursuit movements (Martin 1999). the surface to which an animal’s head is moving (Davies & Green 1994, Lee 1994). Forward placement of the eyes to provide such The function of binocularity in birds symmetrical optic flow-fields in each eye is, how- An explanatory framework for such an interpretation ever, at the expense of comprehensive vision. This of visual field topography is provided by the hypothesis is because the monocular field of a vertebrate eye that, in lateral-eyed vertebrates, binocularity is not is rarely more than 180° in diameter (Martin 1994)

118 G. R. Martin, N. Jarrett & M. Williams

and therefore any forward placement of the eyes of invertebrate recolonization) are reflected in changes inevitably results in a blind area to the rear of the in relative species composition in the diet. We head. Thus, the topography of the complete visual hypothesize that it is these temporal and spatial field is likely to be the result of selective pressures heterogeneities of prey availability that have resulted that have ensured sufficient binocularity for the in the evolution in Blue Ducks of behavioural, control of movement towards targets while at the anatomical and physiological adaptations that same time preserving more comprehensive visual enable the exploitation of a relatively wide range of coverage above and to the rear of the head. The latter invertebrate prey types. However, this broad is clearly beneficial for the detection of predators invertebrate diet would seem to require at least two or for the observation of others in a social group different feeding techniques that are controlled by (Martin 1984, Guillemain et al. 2002). It seems different sensory cues. Thus, Blue Ducks may switch likely therefore that the forward placement of both temporally and spatially between different the eyes in a bird’s skull, which results in a broad primary sense cues depending upon prey type avail- binocular field surrounding the bill, indicates that ability. When exploiting the small sessile chironomid bill position can be brought under visual control and cased caddis fly larvae, which can be scraped during foraging. from rock surfaces, tactile cues from the bill are likely to be the primary cue used to guide behaviour. Hence, the birds can ‘forage blindly’ on, between Frontal binocular field width and and beneath rocks for these prey types. However, foraging underwater visually guided foraging is required when Blue Ducks A particular similarity between the binocular fields are exploiting the larger and more nutritious prey of Blue Ducks and penguins lies in the large absolute (swimming mayfly and active stonefly larvae) that width of their binocular fields. In bird species that are highly mobile within the water column. We forage in air using visual cues to control pecking or suggest that the enhanced binocularity of Blue Ducks, lunging movements towards individual items, the with the visual projection of the bill falling within maximum width of the binocular field is between the binocular field, coupled with their narrow tapering 20° and 30° (Martin & Coetzee 2004). However, in bill, functions to provide the precise visual control amphibious forms binocular field width is broader of bill position necessary for the capture of such than in terrestrial forms and this probably results prey. However, this visually guided foraging by Blue from the fact that upon immersion the widths of the Ducks implies that decreased water clarity is likely monocular and binocular fields are reduced due to to reduce foraging success, particularly for these the loss of refraction by the cornea (Martin & Young larger mobile prey. Thus, rivers that experience 1984, Martin 1999). Thus, a broader binocular field changes in water clarity are likely to be poor locations in air becomes narrower upon immersion. This may in which to concentrate conservation management explain why the binocular field of Blue Ducks (as of this species as any mobile prey that are available measured in air) is broader than in the majority of will be increasingly difficult to catch. Such rivers are terrestrial precision-pecking and lunging birds that those prone to frequent and/or prolonged floods, have been investigated to date (Martin & Coetzee perturbations due to volcanic activity, or those that 2004). drain catchments where land use increases siltation.

Foraging and its sensory bases in The function of bill tip skin flaps Blue Ducks We compared the vision of Blue Ducks and Pink-eared Studies of Blue Duck foraging behaviour and Ducks because they have in common an anatomical diet demonstrate that foraging occurs within specific feature not found in other Anatidae: flexible flaps of microhabitats and that invertebrate prey have hetero- skin that hang from the edges of the upper jaw and geneous spatial and temporal distributions within overhang the lower jaw near the bill tip. The presence the foraging areas (Veltman & Williams 1990, Collier of Herbst’s corpuscles in these flaps suggests a 1991,Veltman et al. 1995, Collier 2004). The pro- sensory function (Gottschaldt 1985) but further nounced temporal variations in the compositions of detailed work on their number and distribution is benthic invertebrate communities (which are mostly required before their possible tactile function in the a consequence of flood events and subsequent patterns foraging of either species can be understood in detail.

Visual ﬁelds and foraging in ducks 119

However, if they have a tactile function such flaps Martin, G.R., Rojas, L.M., Ramirez, Y. & McNeil, R. 2004. The could serve to increase the sensory area on either eyes of oilbirds (Steatornis caripensis): pushing at the limits side of the bill tip within which food items can be of sensitivity. Naturwissenschaften 91: 26–29. Martin, G.R. & Young, S.R. 1984. The eye of the Humboldt detected. This may be particularly important in Blue Penguin, Spheniscus humboldti: visual fields and schematic Ducks as their bill is both narrow and tapered. optics. Proc. R. Soc. Lond. B 223: 197–222. Veltman, C.J. & Williams, M. 1990. Diurnal use of time and We thank Kevin Collier for comments on the initial draft space by breeding blue ducks. Wildfowl 42: 65–86. of this paper. Veltman, C.J., Collier, K.J., Henderson, I.M. & Newton, L. 1995. Foraging ecology of blue ducks Hymenolaimus malacorhynchos on a New Zealand river: implications for conserva- REFERENCES tion. Biol. Conserv. 74: 187–194. BirdLife International. 2006. Species factsheet: Hymenolaimus malacorhynchos. Downloaded from http://www.birdlife.org Received 13 October 2005; revision accepted 20 June 2006. (accessed 14 June 2006). Collier, K.J. 1991. Invertebrate food supplies and diet of blue duck on rivers in two regions of the North Island, New APPENDIX Zealand. NZ J. Ecol. 15: 131–138. Collier, K.J. 2004. Perturbation pathways affecting the avian Procedure lotic predator, blue duck, following a volcanic eruption. Fresh- water Biol. 49: 818–834. Each bird was restrained with the body immobilized Collier, K.J., Moralee, S.J. & Wakelin, M. 1993. Factors affecting and the head position fixed by holding the bill. In the distribution of blue duck Hymenolaimus malacorhynchos on New Zealand rivers. Biol. Conserv. 63: 119–126. both species the bill was held in a specially built Davies, M.N.O. & Green, P.R. 1994. Multiple sources of depth metal holder coated with cured silicone sealant to information: An ecological approach. In Davies, M.N.O. & produce a non-slip surface and the bill held in posi- Green, P.R. (eds) Perception and Motor Control in Birds: an tion by tape (Micropore®). The body was held in a Ecological Approach: 339–356. Berlin: Springer-Verlag. cradle of foam rubber and secured by straps (Velcro®). Gottschaldt, K.M. 1985. Structure and function of avian soma- tosensory receptors. In King, A.S. & McLelland, J. (eds) Form The bill holder was mounted on an adjustable and Function in Birds: 375–461. London: Academic Press. mechanism and the head positioned so that the mid- Guillemain, M., Martin, G.R. & Fritz, H. 2002. Feeding point of a line joining the corneal vertices was at the methods, visual fields and vigilance in dabbling ducks (Anatidae). approximate centre of a visual perimeter apparatus. Funct. Ecol. 16: 522–529. The perimeter’s coordinate system followed con- Kear, J. 2005. Ducks, Geese, and Swans Anseriformes. Oxford: Oxford University Press. ventional latitude and longitude with the equator Kear, J. & Burton, P.J.K. 1971. The food and feeding apparatus aligned vertically in the birds’ median sagittal plane of the Blue Duck Hymenolaimus malacorhynchos. Ibis 113: and this coordinate system is used for the presenta- 483–493. tion of the visual field data (Fig. 2). Each bird’s head Lee, D.N. 1994. An eye or ear for flying. In Davies, M.N.O. & was positioned with the plane through the eyes and Green, P.R. (eds) Perception and Motor Control in Birds: an ° Ecological Approach: 270–291. Berlin: Springer-Verlag. bill tip pointing at an angle of approximately 25 Marchant, S. & Higgins, P.J. 1990. Handbook of Australian, below the horizontal. This head position approx- New Zealand and Antarctic Birds, Vol. 1. Part B. Melbourne: imated that which the birds adopted spontaneously Oxford University Press. when held in the hand. Heads in this position are Martin, G.R. 1984. The visual fields of the tawny owl, Strix aluco L. depicted in Figure 2. The projection of the bill tip Vision Res. 24: 1739–1751. Martin, G.R. 1986. Total panoramic vision in the mallard duck, when measurements were made was determined Anas platyrhynchos. Vision Res. 26: 1303–1306. accurately from photographs and the visual field Martin, G.R. 1994. Visual fields in woodcocks Scolopax rusti- data corrected for this. cola (Scolopacidae; Charadriiformes). J. Comp. Physiol. A. The eyes were examined using an ophthalmo- 174: 787–793. scope mounted on the perimeter arm. The visual Martin, G.R. 1999. Eye structure and foraging in King Penguins Aptenodytes patagonicus. Ibis 141: 444–450. projections of the limits of the frontal retinal visual Martin, G.R. & Coetzee, H.C. 2004. Visual fields in hornbills: field of each eye were determined as a function of precision-grasping and sunshades. Ibis 146: 18–26. elevation (10° intervals) in the median sagittal plane. Martin, G.R., Jarrett, N., Tovey, P. & White, C.R. 2005. Visual To the rear of the head the limits of the retinal visual fields in flamingos: chick-feeding versus filter-feeding. Natur- field were determined at all elevations down to wissenschaften. 92: 351–354. Martin, G.R. & Katzir, G. 1999. Visual field in Short-toed Eagles the horizontal in the Pink-eared Ducks. However, Circaetus gallicus and the function of binocularity in birds. because we did not wish to extend the period that Brain Behav. Evol. 53: 55–66. the Blue Ducks were held the limits of the visual

fields at only one elevation close to the horizontal landmarks within the visual field. The pecten is a highly were determined. From these data (corrected for pigmented structure within the anterior chamber of viewing from a hypothetical viewing point placed the eye that provides nutrition to the retina. It is at infinity) topographical maps of the frontal visual situated above the point where the optic nerve fields and horizontal sections through the visual fields exits. It is highly vascularized and does not contain were constructed. The visual projections of the pectens photoreceptors and hence is a blind area within the were also determined. These provide significant visual field of each eye.