The Jackass Penguin (Spheniscus demersus) as a pelagic predator

Swimming and diving capabilities of Jackass Penguin Spheniscus demersus were examined. Mean maximum speed over a 10 m course in a rock pool was 12.4 km h-' for breeders, 9.5 km h-' for juveniles and 4.6 km h-' for fledglings. Theoretically fledglings could not swim fast enough to catch adult Cape Anchovies Engraulis capensis, the normal prey of adults. Fledglings probably feed on fish larvae. Moulting penguins swam at speeds comparable to fledglings and did not feed. When travelling to the foraging area, breeding penguins swam at 4.8 km h-'. Maximum theoretical foraging range is 24.2 km, but actual range was < 20 km. Mean duration of dive was 22.3 S inside Saldanha Bay and 146 S outside the bay. During long dives outside Saldanha Bay, penguins were probably foraging. At midday, when most Jackass Penguins are at sea, largest numbers were seen outside the bay in 30 to 40 m depth. Penguins fitted with depth gauges dived routinely to 30 m, but spent most time in the upper water layers. Maximum recorded depth was 130 m. There was a positive correlation between distance swum and amount of food ingested. This suggests that the anchovy on which they feed formed small schools which were encountered frequently. INTRODUCTION how the species is adapted to feed on its principal prey, the Cape Anchovy Engraulis capensis (Rand 1960, Penguins are flightless marine birds which rely on Furness & Cooper 1982, Wilson 1985). their swimming abilities to travel to their foraging grounds, and to locate and capture prey. Swimming is much slower than flying so flightlessness would seem MATERIALS AND METHODS disadvantageous. Penguins cannot theoretically cover as large a foraging area as aerial seabirds and are not, Field work was conducted at Saldanha Bay (33"03'S, therefore, adapted to feed on temporally or spatially 17"58'E), southwestern Cape Province, South Africa unpredictable prey where a large area must be covered from May 1980 to July 1981 and in May 1984 particuin order to encounter the prey. Penguins are, neverthelarly at the penguin breeding colony at Marcus Island, less, one of the most successful bird families in the Saldanha Bay. Antarctic and sub-Antarctic, comprising about 80 % of The maximum speeds of Jackass Penguins of differthe bird biomass on many islands (Croxall 1984). ent ages: juveniles (first year birds), feathered chicks Penguins feed on pelagic prey (Zusi 1975) which about to depart to sea, breeding adults, and moulting have several adaptations for avoiding predation, i.e. adults, were determined by timing, with a stop watch, aggregative behaviour (Hamner et al. 1983) which birds swimming over a 10 m distance in a rock pool reduces the probability of predator-prey encounter (dimensions approximately 15 X 2 X 1 m). The birds (Cushing & Harding-Jones 1968), schooling (Pitcher & were given a 1 m stretch in which to accelerate before Partridge 1979, Partridge 1980) where aggregations being timed. They were alarmed when released, apparare highly organised so that the prey can react to ently traveling as fast as possible and generally swimminimize predation after they have been discovered ming directly from 1 end to the other. Any bird that did (Partridge 1980), and die1 vertical migration (Boden & not swim the course directly was omitted from calculaKampa 1967). tions. The state of moult.Was scored as percentage of This paper presents data on the behaviour of the the old feathers lost. Premoult birds were recognised Jackass Penguin Spheniscus demersus to illustrate by their pale plumage and greater mass and postO Inter-ResearchIPrinted in F. R. Germany 220 Mar. Ecol. Prog. Ser. moult birds by their dark plumage, short tail, and lack of skin around the eye (Cooper 1978). Jackass Penguins normally travel by swimming underwater, alternating with periods on the surface. I measured both the surface-paddling speed and the underwater speed of penguins at sea by noting the angular difference between surface 'stops' and the distance between myself and the birds at succesive stops (using a simple device for measuring angular displacement and a rangefinder). The use of vectors coupled with the time taken between surface stops allowed calculation of speed. It was assumed that the penguins travelled in a straight line while underwater. The rangefinder could not be used reliably if the birds were further away than about 300 m. The maximum error in measurements, as determined by calculating unknown distances between randomly placed poles on a beach and then checking with a measuring tape, was less than 10 %. At midday on 28 May 1984, 15 Jackass Penguins guarding chicks were weighed to the nearest 50 g and fitted with autoradiographic speed/distance meters (Wilson & Bain 1984a) which were attached to the feathers with hose-clips (Lishman & Croxall 1983). The speed/distance meter consisted of a spring-mounted bead of radioactive phosphorus 32P and a waterproofed film. The bead position was determined by the speed that the penguin swam. This position was recorded autoradiographically on the film. The time spent travelling at various speeds was ascertained by reading the optical density of the trace on the film (Wilson & Bain 1984a). These birds left the island to forage on 29 May. The chicks were weighed at 1500 h on 29 May and subsequently the nests were checked every hour for returning adults which returned after approximately 11 h at sea. When the penguins with devices returned, they and their chicks were reweighed and the devices removed. The increase in mass of the chicks plus the increase in mass of the adults was assumed to represent a relative measure of the mass of food that the penguins had ingested that day. Penguin density at sea in specific areas was noted by running transects in a 10 m motor-boat at a constant speed (between 25 km h-' and 30 km h-'), and counting the numbers of penguins visible per kilometre. The penguins were spotted from the middle of the boat where the observer was about 4 m above the water. A semi-circle (90" on either side of the boat) was surveyed and any penguins within about 70 m were counted. All transects were conducted on days with good visibility when single penguins on the surface could b e seen easily up to the 70 m limit. Transects were conducted in Langebaan Lagoon, and inside and outside Saldanha Bay (Fig. 1). Fourteen transects were conducted from May 1980 to July 1981 inclusive between 1115 h and 1330 h when most Jackass Penguins are at sea (Wilson 1985). The route of each transect is shown in Fig. 1. Dive durations for travelling and foraging Jackass Penguins were determined with a stop watch by observing the birds from islands (Jutten, Malagas and Marcus) and from stationary boats. The depth to which Jackass Penguins dive and the time spent at each depth was examined by using autoradiographic depth gauges described by Wilson & Bain (1984b). The devices worked on the same principle as the speed/distance meters (see above) with a radioisotope (32P) recording time spent at each depth autoradiographically on film. The device did not record individual dives, but recorded the cumulative total time spent at each depth by the bird per foraging trip. Between 15 December 1980 and 30 May 1981, depth gauges were attached to 15 breeding adults before they left the island at dawn to forage, and recovered when they returned in the evening. In order to investigate the schooling behaviour of the penguins' prey, the stomachs of 61 Cape Anchovies caught by purse-seine fishing boats were examined and graded as empty, full or intermediate. Thirty-one of'the fish were taken from a shoal estimated at l500 t approximately 5 km south of Saldanha Bay on 23 March 1981, and 30 fish were taken from 9 smaller shoals, each estimated between 20 and 50 t, caught approximately 90 km north of Saldanha Bay on 14 April 1981. The stomachs of 122 anchovies taken by 17 Jackass Penguins between 15 March 1981 and 22 June 1981, were graded i n the same way.