Tane 35: 161-173 (1995) OBSERVATIONS ON THE BREEDING OF THE NEW ZEALAND FAIRY TERN by G. Richard Parrish 1 and Gwenda A. Pulham 2 1 14-5 Church Street, Whangarei 2 Unit 2, 1 Parkhill Road, Birkenhead, Auckland 10 SUMMARY Between 1987 and 1994 details were obtained of the breeding ecology of New Zealand fairy tern (Sterna nereis davisae Mathews and Iredale, 1913). Most data were obtained in 1993/94, when 7 pairs were intensively studied. Eggs were laid from mid-november, with re-laying (following the loss of first clutches) occurring up to mid-january. Breeding success was 1.0 chick per pair in 1991/92, 0.83 in 1992/93 and 0.33 in 1993/94. In 2-egg clutches hatching was asynchronous, with one chick at fledging appearing larger than the other. The parents appeared to feed the larger chick by choice, although some competition between siblings was observed. Flooding and predation appeared to be the main factors limiting productivity. INTRODUCTION Information on the breeding ecology of the New Zealand fairy tern (Sterna nereis davisae) is limited. Buller (1888) stated that birds "deposit their eggs on bare shingle, without any attempt at forming a nest" and that "the colour of the eggs harmonised with their surroundings". Oliver (1955) gave details of clutch size, colour and dimension of eggs and a description of nestlings. Guthrie-Smith (1936) described adults surrounding their nest with seashells to match the eggs. Falla et al. (1979), Shaw (1985) and Moon (1988) give the most recent information on nests, egg description, timing of nesting and fledging. This paper gives details of the results of an intensive study of seven fairy tern nesting attempts at Mangawhai (36 06' S, 174 36' E), Waipu (36 S, 174 29' E) and Papakanui Spit on the Kaipara Harbour (36 26' S, 174 12' E) in 1993/94, with further data on all three sites going back to 1987. Nine pairs (two at Waipu, five at Mangawhai and two at Papakanui Spit) made 12 nesting attempts in 1993/94. Data from the other five nesting attempts are also included. 161
METHODS Mangawhai was chosen as the main study area because up to four pairs were known to have nested there in previous years and access to the site was relatively easy. In previous years only one pair had nested at Waipu each year, while two pairs regularly nested at Papakanui Spit. However, access to Papakanui Spit is very difficult and our data indicated that few nests were successful in those years (Parrish & Pulham 1995). Our observations began on 14 November 1993, as the earliest we had ever found eggs was 18 November. The observations continued until 31 January 1994 when all chicks had fledged and left the immediate nesting areas. Observations were made on 40 days, with a total of 142 hours 42 minutes observation. Four nests were closely monitored at Mangawhai (122 hours 34 minutes), one at Waipu (20 hours 23 minutes) and one at Papakanui Spit (11 hours 45 minutes). A second nest at Waipu was observed for 8 hours 54 minutes (concurrent with first nest). Four other nests were monitored (three at Papakanui Spit and one at Mangawhai), but not closely observed. There was one other nesting attempt at Mangawhai, with one egg being laid on 5 December which had disappeared by 7 December (M. Melville pers. comm.). Observations were made at 20-30m distance from the nests with 20x and 25x spotting scopes and 8x binoculars. Egg dimensions were taken with vernier callipers to the nearest 0.01mm and eggs were weighed with 50g pesola scales to the nearest O.lg. The amount of mollusc shell surrounding each nest (up to radius of lm) was calculated from photographs using a planimeter. Information on various aspects of breeding biology collected by us in previous years (1987-93) is also included, as are data from nest record cards held by the Ornithological Society of New Zealand (OSNZ). RESULTS Nest Site Description All nest sites were on low-lying areas of sand and shell above mean high water mark and most were beyond the reach of spring tides. However, all nests were vulnerable to the combination of high tides and waves generated by strong onshore winds. These areas were free of vegetation (except for pioneer seedlings of pingao (Desmoschoenus spiralis) or Spinifex sericeus) and large debris which would restrict vision for the sitting bird. Nest scrapes (n = 24 of 28) were usually situated on slightly raised ridges or mounds of sea shell formed when wind or wave action had eroded the surrounding sand. The shells provided excellent camouflage for eggs, chicks and 162
Fig. 1. New Zealand fairy tern nest scrape and eggs, Mangawhai 24 November 1994. incubating adults (the amount of shell around nests averaged 24.8% of total area, range 13-39%, n = 11) (Fig. 1). The most common shells around the nests were Dosinia anus, Paphies subtriangulata, Struthiolaria papulosa, Glycymeris laticostata, Ponopoa zelandica and Spisula aequilateralis. Strong winds often altered the surrounding substrate during the incubation period. For example, one nest in 1992 was 20m from any bare sand, but before incubation was ended sand encroached to within less than lm of the nest and then retreated again before the chicks were large enough to move. Nest Scrape Construction Birds were seen constructing scrapes two days prior to laying. Trial scrapes were usually formed within several metres of where the nest was later sited, but one pair was seen to engage in digging 100m from where they established their nest two days later. One pair was seen prospecting a patch of shell. Each bird minutely inspected several square metres before commencing to dig. The scrape was formed by a bird lowering itself onto its chest and kicking sand out. The bird then turned several degrees and repeated the digging. Both birds engaged in the scrape construction. One bird then sat in the scrape for up to 20 minutes, 163
while the other bird stood nearby. Two days later these birds laid their eggs in another scrape 2m away. Copulation and Egg-laying Copulation was observed seven times. Before copulation the male stood behind the female and waved his head rhythmically from side to side with neck held erect and wings held partially out and down. The female crouched with her chest on the ground and tail raised; she also swung her head from side to side but not as vigorously. Her wings were also held out, but not at as wide an angle. This pre-copulation display lasted for up to five minutes. The male then mounted and copulation occurred. Copulation lasted for three to five seconds. On six occasions the male stood on the female's back but in one instance a male copulated while maintaining position in flight. On five of seven occasions the male held a fish during the display. The fish was presented three times prior to copulation, once during copulation and once after copulation was completed. We recorded laying of first clutches from 15 November to 17 December. The usual clutch size is two but one egg clutches constituted 22.2% of all clutches. The mean clutch size was 1.77 (n = 54). The second egg was usually laid one or two days after the first egg but in one instance the time difference was four days. The earliest time we observed a pair re-lay after the loss of the first clutch was 12 days. Re-laying of second clutches occurred up to 13 January (1990). Measurements of 13 eggs (3 in 1992 and 10 in 1993, Table 1) ranged from 32.0-39.7mm (x = 34.9mm) x 24.3-26.5mm (x = 25.5mm). Weights of nine eggs ranged from 10.8-15.Og (x = 12.8g). One pair which lost their first clutch laid smaller and lighter eggs when re-laying 21 days later (34.1 x 25.3mm, 11.3g Table 1. New Zealand fairy tern egg dimensions and weights Length (mm) Width (mm) Weight (g) 35.0 26.5 14.5 36.1 25.3 15.0 32.5 25.4 12.0 39.7 24.3 11.0 32.0 25.9 13.5 39.3 25.9 14.5 34.1 25.3 11.3 33.7 25.1 10.8 33.8 25.5 12.4 35.4 25.5-32.8 24.9-34.0 25.9-35.4 24.8 - x = 34.9 x = 25.5 x = 12.8 164
and 33.7 x 25.1mm, 10.8g compared to 36.1 x 25.3mm, 15g and 32.5 x 25.4mm, 12g). Eggs were predominantly ovoid, but varied from ovoid to elliptical ovoid. The base colour was very pale grey tinged with buff or green, overlaid with small irregular markings. These were dark brown on the surface, while those embedded within the shell appeared pale purplish brown to mauve. The markings were often concentrated at the larger end of the shell. Incubation Both parents incubated, but the female sat for longer periods than the male. A total of 53 incubation periods were documented, with the female averaging 58 minutes (S.D. = 27, range 15-125 minutes, n = 22). The male averaged 39 minutes (S.D. = 16, range 5-61 minutes, n = 33). An overnight observation showed that the female sat continuously for at least 10 hours 22 minutes from 20:15 hours (bird already on eggs when observation commenced) till 06:37 hours. As only one night's observation was obtained, we are unable to confirm if the male also incubates at night. A further four incubation periods were documented where the sex of the sitting bird was unknown (range 5-32 minutes). These four observations were early on in the incubation, before we were able to determine the sexes. In one of the nesting attempts the female was colour banded and in another both birds were colour banded. At the other nests we were able to determine sex by small differences in plumage (black at the base of bill and distinct or indistinct demarcation of the leading edge of the black cap). Incubation of the first egg was not continuous for the first two days, with the egg being left unattended for up to 1 hour 45 minutes (range 18-105 minutes, n = 3). This behaviour was also noted in a pair that laid one egg only. One pair commenced full incubation on day three, even though the second egg was not laid until the following day. Observations we made in years prior to this study showed incubation was around 22-23 days, but precise laying dates were not always obtained. Our observations in 1993/94 confirmed hatching at 22-25 days (22-23 days, n = 3, 25 days, n = 1). Hatching was asynchronous on five occasions. In three nests eggs hatched one day apart while hatching was two days apart in the other two nests. Birds have been recorded incubating infertile eggs far in excess of the normal incubation period of 22-25 days. One pair in 1991/92 incubated their two eggs for a total of 37 days. A pair in 1992/93 sat on two eggs for 42 days. These eggs were then removed and replaced with two fertile eggs from Papakanui Spit, which they incubated for a further 13 days (55 days in total), successfully hatching and fledging both chicks. 165
Parental Care of Chicks and Fledging Times Chicks were brooded up to at least six days old. Brooding periods averaged 44.7 minutes (includes three observations when sex of adult was not determined) for both adults (S.D. = 26.1, range 12-87 mins, n = 9), with the female averaging 67 minutes (range 43-87 mins, n = 4) and the male 38 minutes (range 27-49 mins, n = 2). Chicks were left unbrooded for an average of 16 minutes (range 7-32 mins, n = 6) on days five and six. By day seven the chicks were not brooded at all. The chicks left the nest scrape at two to three days old. They then moved to another scrape close by (<5m away). From six days old the chicks became progressively more mobile and moved up to several hundred metres away from the original nest site seeking shelter under flotsam or sat in a scrape they had dug themselves. At least one parent remained on guard in the immediate vicinity of the chicks and chicks were not left totally unattended for any length of time until 14-15 days old. When adults were alarmed by the presence of potential predators the chicks froze. During a storm, a chick at Waipu in 1993/94 was led from the original nest site on the exposed seaward side to the sheltered inland side of the sandspit and then back again once the storm had subsided. At a one chick nest where the chick was five to six days old it was fed at an average rate of one fish per 117 minutes (n = 6) with periods of up to 135 minutes when no food was consumed by the chick. During brooding, the male offered fish either to the chick or to the female. She in turn either fed the chick or ate the fish herself. On the occasions when both the chick and the female declined the offered fish, the male ate it himself. However, when the male was brooding, the female offered food only to the chick and if it was declined she ate it herself. In a two chick clutch when the chicks were nine to ten days old, there was some competition for food. When both chicks ran towards the incoming parent bringing food, the larger chick usually arrived first and obtained the food. At 14-15 days old they were fed five and four times in 132 minutes (average: once every 26 and 33 mins respectively). When 22-23 days old, these chicks were fed 18 times in a six hour period, with the larger chick being fed 12 times and the smaller one six times. The number of feeds in each hour of the six hour observation was 2, 0, 1, 2, 4, 9. During this observation period the chicks were very mobile, moving over an area of about 2ha. The chicks were usually apart and at this age we saw no evidence of the larger chick ousting the smaller when food was offered. The parents continued to feed chicks after fledging, with one chick observed being fed at 35 days and another begging at 50 days old. In a nest containing one chick and one egg, the female left the nest in response to the incoming male so abruptly that she knocked the chick over on three occasions. On one of these occasions the chick was knocked onto its back and it took several 166
Fig. 2. New Zealand fairy tern nestlings 5 days old, Waipu January 1991. Note rear chick is larger than front one. seconds to right itself. Fledging occurred at 22-23 days old. By day 21 some chicks were capable of flights of 2-3m, but by day 23 they were able to fly up to 100m. Description of Chicks The down of nestlings was pale grey above, tipped with buff and lightly marked with brown/black, giving a sandy appearance. Below they were very pale grey. The bill and legs were reddish pink (Fig. 2). Within days of brooding ceasing, emerging feathers on the back, mantle and wing coverts were visible and by 10-12 days these were mottled cinnamon/orange. The rest of the body was still downy. The bill and legs darkened to become reddish brown and dull orange respectively. At 20-22 days the mantle and back feathers were grey (still some down showing) with bands of pale cinnamon and dark brown. The upper tail coverts were grey tipped with pale orange and brown. The tail was white tipped with pale orange. Underparts including tail, were white. Primaries were dark grey with white shafts. The two outer primaries had blackish vanes with white edges. 167
Secondaries were grey, inner edges white. Lower coverts pale grey with faint white edges and bands of pale orange. Upper coverts pale grey, leading edge of wing white. The crown had cinnamon feathers which became blackish with cinnamon tips towards the hindcrown and nape. A pale grey/white superciliaty stripe ran from the lower forehead to nape. There was a dark rust band running from eye to bill with a distinct semi-circle of black feathers in front of the eye. The bill was dull brown with a bright yellow gape while the feet and legs were dull orange with black claws. The eye was dark brown (Fig. 3). Food Items In 1993, a chick regurgitated a fish which was identified as a goby {Favonigobius lentiginosus). In this study we observed the feeding of fish to females and chicks on 184 occasions. We were able to identify 10 items: 7 elvers (Anguilla sp.) and 3 flounder (Rhombosolea sp.). As the viewing distance was usually in excess of 20m and the exchange of food so quick, it was usually impossible to identify the items. However, on 55 occasions the item was seen and their tentative identifications were discussed with B. Stephenson, Curator of Fishes at the Auckland Institute and Museum, and R. McDowell, National Institute of Water and Atmospheric Research, Christchurch. From these discussions we tentatively identified the following: juvenile spotty {Notolabrus celidotus) or triplefin {Forsterygion sp.) (n = 12), small blackish fish, possibly goby or gudgeon (Grahamichthys sp.) (n = 25), sprat-like fish, probably yelloweyed mullet {Aldrichetta forsteri) or possibly smelt {Retropinna retropinna) (n = 15) and translucent fish possibly whitebait {Galaxias sp.) or juvenile smelt (n = 3). We most commonly observed the adults feeding within the estuaries, but on occasions, they fed at sea just beyond the breaking waves. Once the chicks were mobile, they often pecked at items they encountered on the beach. On three occasions we saw chicks pecking at seaweed and apparently removing and swallowing the small inflation pods. On other occasions they were seen to pick small animals from shells, driftwood and seaweed. These animals were either insects, isopods or amphipods. Other Adult Behaviour at the Nest Incubation Changeovers The relieving bird usually landed l-4m from the sitting bird. Most times the incoming bird called. Generally, the relieved bird left without hesitation (n = 48), but was sometimes reluctant to leave (n = 18). The reluctant bird frequently indulged in rearranging shells up to 5m away from the nest. Females were usually the reluctant birds, but males also exhibited this behaviour. On one 168
Fig. 3. New Zealand fairy tern chick aged 18 days, Mangawhai 11 January 1992. occasion a female was seen to rearrange shells immediately around the nest while still sitting on the eggs At one nest, the male was seen to attempt to incubate a hatched egg shell lying some 2m from the scrape. After trying to align the half shell into a position for sitting on, the bird abandoned it and settled itself upon the recently hatched chick and remaining egg. Egg Turning Turning of the eggs was recorded 16 times. Egg turning consisted of the sitting bird rotating through 360 while shuffling the body. This was sometimes accompanied by the bird prodding the eggs with its bill. On one occasion a female turned through 360 three times in succession to realign the eggs. The time between egg turning averaged 67 minutes (range 10-145 minutes, n = 9). Sleeping, Preening and other Behaviour During long incubation periods the birds, particularly the female often appeared asleep. Preening frequently took place after a bird had left the nest or after hunting, but occasionally birds would preen while incubating. 169
On days when the wind speed was less than 15 knots, the incubating bird positioned itself without any particular orientation. However, on windy days, the birds always faced directly into the wind. In very strong winds, sand would accumulate around the birds. They removed this by fluffing out their feathers and vigorous shaking. If sand continued to fill the nest scrape, the birds kicked it out in the same manner as when they excavated the scrape. Intra- and Inter-specific Interactions There was a marked difference between the sexes in their reactions to intruders. The female usually reacted to invasion of the territory by crouching lower to the ground and remaining motionless while the male would readily leave the nest to intercept and drive off the intruder. A total of 73 interactions between fairy terns and other species were recorded. Territorial disputes with other fairy terns were observed on 23 occasions. These usually occurred between an incubating pair or a pair with chicks, and pairs that had not commenced nesting or had lost a nest. It usually involved the resident pair aerially chasing the invading pair from the territory. However, on several occasions non-breeding birds were tolerated within 1 m of the nest without any response from the sitting bird. Fairy terns interacted with variable oystercatcher {Haematopus unicolor) n = 7, New Zealand dotterel {Charadrius obscurus) n = 7, Eastern bar-tailed godwit (Limosa lapponica baueri) n = 1, red-billed gull {Lams novaehollandiae scopulinus) n = 4, southern black-backed gull (L. dominicanus) n = 23, Caspian tern {Sterna caspia) n = 1, white-fronted tern {S. striata) n = 4, rock pigeon {Columba livia) n = 1 and people n = 2. Interactions usually involved aerial chases to beyond the territory boundary, but with variable oystercatchers and New Zealand dotterels the fairy terns usually swooped low over the intruder, causing them to duck to avoid being struck. Both these species share nesting habitat with fairy terns. Often there was no response from the terns to New Zealand dotterels, even when they approached very close to the nest, but they reacted much more vigorously when oystercatchers were nearby. In one instance three variable oystercatchers were seen to attack a fairy tern nest. While a nesting pair of fairy terns were being distracted by two oystercatchers a third oystercatcher approached the nest and pecked at the nest scrape (J. Parrish pers. comm.). At this point the observer intervened and prevented any further disturbance. On one occasion, 120 godwit, c. 200 lesser knot {Calidris canutus) and 17 turnstone {Arenaria interpres) congregated at high tide close to a nesting pair of fairy tern. The numbers increased as high tide approached and the terns became increasingly more nervous, eventually responding by attacking the birds, 170
particularly the godwits, which were forced to move metres away. Productivity and Causes of Nest Loss Sixteen of 34 nests we have observed over the past seven years failed to produce fledged chicks. The most common cause of nest loss is flooding (n = 9). This is particularly the case at Papakanui Spit (n = 7), which is subject to frequent flooding because of its low-lying nature. To our knowledge, no chicks have fledged at Papakanui Spit in the last four years. Flooding at Waipu and Mangawhai is not so frequent and we know of only one nest at each site that has been lost to floods in the past seven years. Eggs or chicks disappeared from nests on six occasions. We believe predation to be the cause of this disappearance. The predator involved in four of these losses have not been identified. However, in 1993/94, two nests at Mangawhai were destroyed, most likely by variable oystercatchers. In a nest of one chick and one egg, the chick and egg were stabbed but neither the chick nor the egg contents were devoured. This was also the case in the other nest of one egg. As stated above, three oystercatchers were seen interfering with a nest at Mangawhai. Similar behaviour was also observed by one of us (GAP) at Waipu where a chick was attacked away from the scrape. Although there were no witnesses to the actual predation of the eggs and chick, we believe the documented attacks by oystercatchers and the evidence left behind (egg contents and chick not eaten or removed), strongly suggests that they were the culprits. Wardens monitoring nesting shorebirds at Mangawhai and Waipu have recorded or suspected infertility in fairy terns. All unhatched eggs (1991-94) were opened by GRP to determine cause of failure to hatch. One pair had two infertile eggs in 1991/92 at Mangawhai and a pair in 1992/93 also had infertile eggs again at Mangawhai. In 1993/94, three nests contained one infertile egg each, one at Waipu and two at Mangawhai. Two of these pairs were nesting for the first time, while the age of the other pairs was unknown. In the 1991/92 breeding season, six nesting pairs fledged six chicks (3 nests each with two fledglings). In 1992/93 six pairs produced five chicks (3 nests each with one and one with 2 fledglings); this included two chicks which fledged from transferred eggs from Papakanui Spit. In 1993/94 nine pairs made 12 nesting attempts hatching 5 chicks and fledging three (clutches included 7 of one egg and 5 of two). DISCUSSION Shaw (1985) and Falla et al. (1979) stated that the New Zealand fairy tern usually breeds in isolation, in contrast to the colonial nesting of the Australian 171
subspecies. We have observed nests within 20m of each other while G. Moon (pers. comm.) found three nests within a radius of 8m. This suggests that the New Zealand subspecies might nest colonially, i.e. in loose congregations, if there were sufficient pairs and habitat. Moon (1988) recorded birds forming nest scrapes several weeks before egg laying. Although we observed courtship display up to two months before nesting, we did not observe scrape construction until two days before egg laying. However, we did not conduct many observations in the pre-nesting period. Shaw (1985) stated that on completion of copulation the female flies off with the fish with the male in pursuit. In all our observations the female remained on the ground for several minutes after copulation and it was the male that flew off first. Shaw (1985) stated that egg laying is in November while Moon (1988) gives early December as the egg laying period. The earliest date for laying we are aware of is 15 November (1994), while the latest date for a first clutch is 17 December (1992). Re-laying has been recorded up to 13 January. This is a much shorter breeding season than other shore nesting birds in New Zealand. The New Zealand dotterel will lay eggs from late August and following re-laying can still be fledging chicks in mid March. The variable oystercatcher commences egg laying a month or so later than New Zealand dotterel, but chick rearing also extends through to mid March. In our experience New Zealand fairy terns will re-lay if their first eggs are lost. They do not appear to re-lay if the second clutch is lost. Guthrie-Smith (1936) recorded birds surrounding the nest with coloured shells that matched the colour of the eggs but we did not record this behaviour. Shaw (1985) gave the incubation period as 18 days, while Moon (1988) stated 21 days. T. Shaw (pers. comm.) in 1990 watched two eggs being laid and the two chicks hatching 22 days later. We recorded incubation periods of 22-25 days. Incubation was usually around 22-23 days, but full incubation did not start until the full clutch was laid which can be up to four days after the first egg. Even birds with only one egg appeared to wait a couple of days before full incubation commenced. Some authors (Oliver 1955, Shaw 1985) stated that fairy terns feed primarily on small fish, but were unable to identify which species. Moon & Ell (1979) recorded anchovies {Engraulis australis) being fed to chicks. Shaw (1985) recorded mainly small fish with some gastropods, crustaceans and plants. This appears to be a quote from Hitchcock (1959) who examined the stomach contents of six 5". n. nereis. In this study we were able to positively identify three species of fish fed to females and chicks. In the other 55 observations, the identification must be regarded as tentative because of the distance of the observations. We also observed chicks eating either insects, isopods or amphipods and apparently 172
consuming inflation pods of seaweed. We were unable to confirm gastropods or crustaceans as reported by Shaw (1985) or Hitchcock (1959). Few data have been published on S. n. davisae in New Zealand. It is tempting to make comparisons with the Australian populations, but we feel that we should leave this to others more qualified and to a time when there are enough data to justify making more reliable comparisons than are possible at the moment. ACKNOWLEDGEMENTS This paper is dedicated to the memory of Jan Parrish (1946-95), a lover of fairy terns. Numerous people have assisted us over the years, in particular we would like to thank Frances and Morrie Crawford for the use of their boat, Simon Chamberlin, Peter Penny, Bill Walsh and Butch and Venus Woodcock. We also thank the wardens Jason Roxburgh, Tim Shaw, Noel Henry, Ross Sinclair, Peter Graham, Demien Reed and Mark Melville for their assistance. The Ornithological Society of New Zealand provided a grant of $450 to assist us in our study. Fred Brook identified species of seaweed and shells. Greg Sherley without whose help this paper would not have been possible. Judy Roberts typed the manuscript. REFERENCES Buller, W.L. 1888: "A history of the birds of New Zealand. 2 volumes, 2nd ed." Published by the author, London. Falla, R.A, Sibson, R.B. & Turbott, E.G. 1979: "The new guide to the birds of New Zealand." Collins, Auckland. 247pp. Guthrie-Smith, H. 1936: "Sorrows and joys of a New Zealand Naturalist." A.H. and A.W. Reed, Wellington. 252pp.. Hitchcock, W.B. 1959: A review of the 'least' terns in Australian waters. South Australian Ornithologist 22: 87-106. Mathews, G.M. & Iredale, T. 1913: A reference list of the birds of New Zealand. Ibis S10: 245. Moon, G. 1988: "What New Zealand Bird is that?" Weldon, Sydney. Moon, G. & Ell, G. 1979: "The birds around us." Heinemann, Auckland. 207pp. Oliver, W.R.B. 1955: "New Zealand Birds." A.H. & A.W. Reed, Wellington. 661pp. Parrish, G.R. & Pulham, G.A. 1995: Population size, productivity and post-breeding movements of the New Zealand fairy tern. Tane 35: 175-181. Shaw, P.W. 1985: Fairy Tern In: "Reader's Digest Complete Book of New Zealand Birds." C.J.R. Robertson (ed.). Reader's Digest, Sydney 319pp. 173
174