THE DEMOGRAPHY OF NEW ZEALAND'S CYGNUS ATRATUS POPULATION M W ILLIAM S. Introduction. Population dispersion, size and structure

Transcription

1 THE DEMOGRAPHY OF NEW ZEALAND'S CYGNUS ATRATUS POPULATION M W ILLIAM S Introduction Cygnus atratus is a conspicuous inhabitant o f the lakes, lagoons and estuaries o f New Zealand. Native to Australia, it was first introduced to New Zealand in 1864 (Thompson 1922) but, coincidental w ith this, natural im m igration seems also to have occurred (K irk 1895). The species acclimatized very rapidly, exploiting a niche unoccupied since the extinction, several centuries previously, o f the closely related Cygnus sumnerensis, and by 1900 breeding populations were established on all m ajor freshwater lakes in both N orth and South Island (Thompson 1926) and on Chatham Island (Richards 1950). Throughout almost all o f its history in New Zealand, C. atratus has been exploited as a game-bird and its eggs have been collected. Because o f its recreational value and because, in some areas, it is becoming an agricultural pest, current management practices include biannual m onitoring o f the national population by means o f aerial surveys and the annual banding o f about 4000 m oulting adults and cygnets to determine dispersal, survival and levels o f exploitation. Population dispersion, size and structure W ithin New Zealand, swans prefer the large permanent bodies o f fresh water, avoiding small impoundments and flow ing water. They are quick to e xplo it the food available in tem porarily flooded swamplands and farmland, and they occur throughout the year on most coastal lagoons and large estuaries. A ll o f the principal habitats are sampled during the biannual aerial surveys and estimates of the birds present are made. Surveys in November record the dispersion of the population at the height o f the breeding season, w hile those made in January distinguish between cygnets and birds in adult plumage and so provide an estimate o f that season's production. The dispersion o f the population in January 1979 is shown in Fig 1. The population on the New Zealand mainland at that tim e was estimated to be , o f which approxim ately 5000 were fledged or near-fledged cygnets. Swans on Chatham Island, 800 km east o f New Zealand, are not regularly surveyed. The most recent estimate is approxim ately 3000 birds (in 1978, E S Bucknell pers comm), although previous estimates there were as high as (Lindsay et al 1959, Internal A ffairs Dept files). Estimates o f the number o f breeding pairs at most breeding areas have been made 147

2 Fig 1. Principal concentrations of Cygnus atratus in New Zealand in January 1979, showing the proportion of the population at each locality which bred (black sector), an additional proportion which in the past three years has also bred (stippled sector) and the non-breeding fraction (open). 148

3 (Table 2) and the 1978/79 national breeding population was about 4600 pairs, approximately 17% o f the national adult population, However, the number o f pairs nesting annually varies and this is incorporated diagrammatically into Fig 1 and discussed further in the section on 'Population characteristics'. Breeding does not occur on any o f the marine habitats but swans are present there at all times o f the year. This suggests that there may be tw o separate elements w ith in the population w hich contribute to the distribution observed during each breeding season adults w hich attem pt breeding and non-breeders, the latter perhaps a com bination o f birds to o young to breed plus adults o f breeding age w hich make no nesting attem pt. The ages o f birds present at one marine habitat (Farewell Spit) towards the end o f the breeding season have been investigated; during the period 1976 to 1980, 153 swans originally banded as cygnets were caught during th eir wing m oult; 70 (46%) were one year old, 47 (30%) between one and three years o f age (it is thought that most swans first commence breeding when fo u r years old), 32 (21%) between fo u r and eight years old and 4 (3%) older than eight years. Perhaps the marine and estuarine inhabitants are mainly pre-breedérs. Non-breeders seem also to comprise the major proportion o f the swans present on the principal breeding areas. A t all nesting localities where the numbers o f nests in colonies were counted (Table 1), a maximum o f one-third o f the swans present Table 1. The number of present in breeding areas nests counted and tht Breeding area Number Number of and date o f nests swans present Lake Ellesmere Lake Ellesmere Lake Emma Lake Heron O karito Lagoon Vernon Lagoon Lake Waihola % total swans breeding seem to have participated in nesting. Censuses o f broods at three areas (Lake Emma 1973, Lake Waihola 1971 and Lake Whangape 1975) recorded that 15% to 27% o f the swans present were attending broods. The ages of swans present at these principal breeding sites during a breeding season have been estimated only fo r Lake Ellesmere. There, at the end o f 1977, 70% o f the 9300 adults present were estimated to be older than fo u r years (Williams 1979). Are the birds present at these principal breeding areas m ostly the older birds? If so, w hy do so few attem pt to breed? 149

4 Population dispersal Should New Zealand's population o f C. atratus be managed on a national basis or is there evidence that the population is sub-divided regionally? A ttem pts to answer this question have in itia lly involved determining the lim its of dispersal o f swans reared at major breeding localities and examining the extent to which these areas o f dispersal overlap. Inform ation about movements of swans from Lake Ellesmere and Waikato lakes has been derived from the analysis o f bands returned by hunters (Williams 1977). Most swans banded as cygnets on Lake Ellesmere (Fig 1) were recovered either at the lake itself or from coastal wetlands to the south. Although some birds were shot in N orth Island there was no evidence o f major movements in th at direction. Hunting ceased in South Island in 1974 but aerial surveys confirm that movement between Lake Ellesmere and all points south to Invercargill Estuary continues. Waikato swans (Fig 1) were also recovered in a clearly defined region. A lm ost 90% o f the band recoveries were made w ith in 80 km radius o f the banding site and most o f the more distant recoveries were from Northland. Aerial surveys confirm very large movements o f swans from the W aikato lakes to the enclosed harbours in Northland. These findings tell as much about the dispersion of hunters as about that o f swans, but to the w ild life manager this is extremely useful. However, hunting e f swans is no longer perm itted w idely throughout the country. To describe more completely the dispersion o f swans both during the hunting season and at other times o f the year, banding and collaring programmes commenced in 1974 at four localities. The results o f these studies (Williams in prep) may be briefly summarized: Lake Wairarapa: 500 cygnets banded and collared annually fo r five years. (i) Collar sightings: approxim ately Those reported away from the lake were m ainly at Farewell Spit, in Marlborough or on lakes in Manawatu. Small numbers were regularly seen at Lake Ellesmere but sightings elsewhere were rare. (ii) Band returns: almost 90% o f 355 were made at Lake Wairarapa, the remainder in Marlborough and Hawkes Bay (an area where few collar sightings were made). Thus, Lake Wairarapa swans remain m ostly in the area o f Farewell Spit, Marlborough and the lake itself w ith some birds regularly reaching Hawkes Bay and Manawatu. Hawkes Bay: 100 to 150 cygnets banded and collared annually fo r five years. (i) Collar sightings: 120, 50% w ith in the Hawkes Bay d istrict and a further 20% at Lake Wairarapa. Other sightings were scattered throughout Manawatu, 150

5 Marlborough and at Farewell Spit. No sightings were made in the northern half of N orth Island and only fo u r south o f the Marlborough district. (ii) Band returns: 56, 15 o f which come from beyond Hawkes Bay. Two swans were shot at Rotorua where no collared birds have been reported, and 10 from Lake Wairarapa and Marlborough. This small population shows only lim ited dispersal beyond its natal area, most o f this south o f Lake Wairarapa. Rotorua lakes: 686 cygnets banded and collared over four years. (i) Collar sightings: 300, 70% of which were in the immediate area of the Rotorua lakes and Lake Taupo. Most other sightings were from the Waikato area but Rotorua swans have been seen in Manawatu, at Farewell Spit, on Lake Wairarapa and Lake Ellesmere and in Otago. (ii) Band returns: 48, 10 o f which came from areas other than the Rotorua lakes and Lake Taupo. Four birds were shot on the Waikato lakes, three in the Hawkes Bay d istrict and three on coastal lagoons of the Bay of Plenty. No banded birds have been shot on Tauranga Harbour. Dispersal characteristics o f this population are not yet clear. Considerable movement to the Waikato wetlands seems to occur, especially in w inter when water levels in the Rotorua lakes are high. However, numerous swans, sometimes in excess of 1500, are present throughout most o f the year on nearby Tauranga Harbour. Their origin is unknown but it is assumed that they may comprise birds from both the W aikato and Rotorua populations. Farewell S pit: This area, a designated wetland o f international importance, is the major m oulting site fo r swans in New Zealand and, at the peak o f the m oult, approxim ately 15% of New Zealand's C. atratus are present. Alm ost 4500 m oulting swans have been banded and collared over five years. (i) Collar sightings: almost 1200, scattered from Northland to Otago. Most sightings have been from Marlborough, Manawatu and Lake Wairarapa. Collared swans have been observed breeding in these localities and also in Hawkes Bay, Rotorua and at Lake Ellesmere. (ii) Band returns: 204, the m ajority from Lake Wairarapa and Marlborough but also from Hawkes Bay and Waikato. (iii) Recapture o f banded birds: 153 swans, banded as cygnets at various localities, have been recaptured at Farewell Spit, 90 (59%) from Lake Wairarapa, 28 (18%) from O karito Lagoon, 14 (9%) from Waikato lakes, 10 (7%) from Lake Ellesmere 151

6 Fig 2. The geographic sub-division and size of New Zealand's mainland Cygnus atratus population and the pattern of major (thick arrow) and minor (thin arrow) movements between sub-populations. 152

7 and 11 (7%) from Hawkes Bay. No birds banded on the Rotorua lakes (the only other banding site) have been caught although sightings o f Rotorua collared swans have been made. Thus, swans present at Farewell S pit during the peak o f m oulting are drawn from a wide area but predom inantly from nearby breeding areas in Marlborough and Lake Wairarapa, from O karito Lagoon and probably Manawatu. O karito Lagoon: A pproxim ately 100 cygnets have been banded annually since 1973 but not marked w ith collars. A ll but fo u r o f 51 band returns have come from the West Coast area and, apart from the presence o f birds m oulting at Farewell Spit, it seems the O karito population is a sedentary one. These data on movements and the changes in numbers and distribution revealed during the twice-yearly aerial surveys indicate that there are several discrete or partially discrete populations o f C. atratus in New Zealand, as illustrated in Fig 2. W ithout doubt, some o f these populations overlap, sharing a common w inter feeding area (eg Ellesmere and Otago/Southland, Waikato and Rotorua) but there is no evidence o f widespread interm ingling throughout the year nor o f interbreeding consistent w ith the hypothesis o f a mobile national population. Population characteristics In Table 2 principal characteristics o f the various regional swan populations are summarized. Nesting m ethod C. atratus is popularly considered to be a colonial nesting species (Kear 1972). However, in many situations in New Zealand it is classically te rritoria l, each breeding pair defending several hectares o f open water (to which they confine all o f th e ir feeding activities) and swamp edge (where the nest is constructed). This te rrito ry may be established several months prior to laying, is usually vigorously maintained throughout the fu ll period o f the cygnets' development, and is vacated only briefly during the period o f the wing m oult. Solitary nesting and the establishment o f a te rrito ry occurs on: (a) (b) (c) (d) small water bodies (5 to 20 ha); large water bodies ( > 20 ha) where non-breeding birds do not permanently reside throughout the breeding season; large water bodies where aquatic macrophytes are distributed in a thin band around the margins and not distributed throughout; areas where the water level is constant or almost so throughout the w inter and spring period. 153

8 By contrast, colonial nesting, characterized by some nests being spaced a mere pecking distance apart and defence being restricted to the immediate area o f the nest, is usually the nesting method adopted at: (a) (b) (c) (d) (e) large water areas; areas where large numbers of non-breeders are present throughout the breeding season; areas where water levels fluctuate throughout the w inter and spring period and experience considerable summer drawdown; areas where food is patchily distributed or is distributed all over the wetland in both circumstances being extremely abundant; estuarine areas. Colonial nesting appears to be, as Kear (1972) has suggested, an adaptation which allows abundant food in areas o f unstable water levels to be exploited. Tim ing and length o f breeding season On those wetlands w ith stable w inter and spring water levels, where most solitary nesting occurs, the tim ing of nesting is regular, commencing in July and extended to early October. The tim ing of breeding o f colonial nesters is irregular and is determined by water levels. Colonial nesting occurs as levels decline from the w inter peak and seldom commences before September. However, high water levels during spring may delay any nesting attem pt until November and, in some cases, into the new year. Once commenced, however, nesting in colonies is highly synchronous and more than three weeks seldom elapse between the first and last egg in a colony. There are numerous examples of fluctuating water levels inhibiting breeding. A t Lake Whangape in 1978 water levels gradually declined from the w inter peak in July and early nesting commenced in late September as traditional nesting areas became exposed. However, in late October levels increased again and remained higl, throughout November. Most of the adult male swans sampled in October, November and December o f this year showed testicular regression and when water levels declined throughout December and January to expose the usual colonial nesting sites, few pairs were able to respond and nest. A t Waituna Lagoon in 1973, the o u tle t to the sea became blocked in September and high water levels remained throughout the summer. In February 1974, however, the blockage was breached and, once the water level declined, nesting commenced, approxim ately five months later than usual. These tw o examples indicate that in New Zealand C. atratus is a strictly seasonal 154

9 breeder but is able to delay the onset o f nesting until water levels allow access to suitable nesting sites and food becomes readily available. Variations in water levels during the breeding season can stop all nesting activity and prevent further breeding attempts th at season. Clutch size An extensive study at Lake Ellesmere almost 20 years ago (Miers and Williams 1969) recorded the mean clutch as 5.4, the most common clutch being 5 eggs. The clutch sizes at Ellesmere have been recorded every year since 1975, each year's sample comprising at least 500 nests. There has been a progressive decline, the mean clutch being 5.0, 4.9, 4.0, 4.3, 4.0 in 1975 to 1979 respectively, reflecting the deteriorating state o f Lake Ellesmere as swan habitat (Williams 1979). Clutch size data are available from fo u r other areas, all colonial nesting situations: the mean sizes o f clutches were 4.9 at O karito Lagoon (23 nests in 1973), 5.2 at Lake Wairarapa (70 nests in 1979), 5.1 at Lake Whangape (538 nests in 1974) and 5.7 at Vernon Lagoon (225 nests in 1975). In all cases, the most common clutch was o f five eggs. There are too few data to determine what differences may occur (a) between nesting areas, (b) at the same nesting area between years and (c) between solitary and colonial nesters. Brood rearing Cygnets are reared either in fam ily broods, guarded and attended by their parents, or they may be reared in creches which appear to be attended by only one pair o f adults. Family broods occur in all situations where pairs nest solitarily, and, where colonial nesting occurs, at least some o f the broods are raised by their parents. W ork at Lake Ellesmere in 1976 which involved tagging entire broods o f cygnets at hatching showed that 32% o f the cygnets were reared in fam ily units; the remainder were distributed in creches o f various sizes. One extreme example was a creche o f 40 w hich contained 19 tagged cygnets, derived from 15 different broods. If the untagged cygnets were sim ilarly derived, this creche may have contained cygnets from up to 30 different broods. However, approxim ately 70% o f the creches handled contained cygnets derived from tw o, three or four broods (Williams unpubl data). A t Lake Ellesmere, broods reared as fam ily units were m ostly those hatched early and taken by their parents to the more remote parts of the lake. Creching occurred when large numbers o f broods attempted to feed simultaneously in the same confined area. The patchy distribution o f food there forced broods into close contact and the apparent lack o f aggression between guardian adults allowed frequent interchange o f cygnets between broods. 155

10 Table 2. The demographic characteristics o f the various sub-populations o f C. atratus on the N orth and South Islands o f New Zealand. L ittle is known o f the Chatham Island population. It numbers 3000 swans and is believed not to move beyond the Chatham group o f islands. P rin c ip a l b re e d in g area N O R T H L A N D / W A I K A T O L a k e W h a n g a p e, W a h i a n d o th e r s c a tte re d la k e s in W a ik a to b a s in a ll fre s h w a te r. F e w o n N o r th la n d fr e s h w a te r w e tla n d s. R O T O R U A W id e ly s c a tte re d o n a ll la k e s b u t m a in ly L a k e s R o to r u a a nd R o to e h u a ll fre s h w a te r H A W K E S B A Y S m a ll n u m b e rs o n a ll fre s h w a te r w e tla n d s M A N A W A T U S m a ll n u m b e rs o n a ll fr e s h w a te r w e tla n d s W A I R A R A P A / M A R L B O R O U G H L a k e W a ira ra p a (fr e s h w a te r ) V e r n o n L a g o o n (fr e s h w a te r ) M a in lo c a tio n o f n o n -b re e d e rs in b re e d in g season L a k e W h a n g a p e (fr e s h w a te r ), K a ip a ra a n d P a re n g a re n g a H a rb o u rs (e s tu a rin e /m a r in e ) T a u ra n g a H a r b o u r (m a r in e ). L a k e T a u p o (fre s h w a te r) S m a ll n u m b e rs w id e ly scattered. A ll fre sh w a te r. P o s s ib ly s o m e m o v e m e n t s o u th to L a k e W a ira ra p a W id e ly s c a tte re d o n a ll fr e s h w a te r w e t la n d s. C o n c e n tr a tio n s o n L a k e s H o ro w h e n u a and P apaitonga to w a rd e n d o f season A t m a in b r e e d in g a re a s p lu s m a rin e are as a t H a v e lo c k a n d F a re w e ll S p it B re e d in g season V a r ia b le o n L a k e W h a n g a p e b u t seem s to h ave t w o p e a k s, o n e in A u g u s t/s e p te m b e r f o llo w in g in it ia l w a te r le v e l d e c lin e f r o m w in t e r p eak a n d a n o th e r in N o v e m b e r t o J a n u a ry fo llo w in g f u r th e r d e c lin e w h ic h e x p o s e s is la n d s in la k e. E ls e w h e re re g u la r S e p te m b e r to N o v e m b e r A u g u s t t o O c to b e r re g u la r A u g u s t t o O c to b e r re g u la r J u ly t o O c to b e r re g u la r L a k e W a ira ra p a o n s e t is ir re g u la r b u t m o s t ly A u g u s t t o O c to b e r. V e r n o n L a g o o n ir re g u la r, m a in ly J u ly t o S e p te m b e r N e s tin g m e th o d E a r ly n e s tin g a t L a k e S o lita r y W h a n g a p e a n d e lsew h e r e s o lita r y. L a te n e s tin g a t L a k e W h a n g a p e is c o lo n ia l C lu tc h sizes L a k e W h a n g a p e , 5.1 in c o lo n ia l n ests. N o re c o rd s f o r s o lita r y n e s te rs S o lit a r y S o lita r y L a k e W a ira ra p a b o th s o lita r y a n d c o lo n ia l. V e r n o n L a g o o n c o lo n ia L a k e W a ira ra p a , 5.2 in 7 0 c o lo n ia l n ests. V e r n o n L a g o o n , 5.7 in c o lo n ia l n e sts B ro o d re a rin g C y g n e t p r o d u c tio n F a m ily b ro o d s. N o o b s e rv a tio n s o f c re c h e s : range in past fiv e years F a m ily b r o o d s : ra n g e in p a s t f o u r y e a rs F a m ily b ro o d s : R e c e n t c o u n ts «a b o u t F a m ily b ro o d s : A n n u a l in c re a s e o f o v e r p a s t th r e e y e a rs L a k e W a ira ra p a b r o o d s re a re d o n m a in w a te r b o d ie s re m a in in fa m ilie s. T h o s e reared in pastu re o r on ephem e ra l w a te r s c re c h e. V e r n o n L a g o o n b o th f a m ily b r o o d s a n d c re c h e s L a k e W a ira ra p a : a p p r o x im a te ly b u t e x te n d e d b re e d in g m a d e e s tim a te v e ry d i f f i c u l t. A v e ra g e a n n u a l p r o d u c t io n p r o b a b ly w it h in ra n ge V e r n o n L a g o o n : H ig h ly v a ria b le n il in past fiv e years

11 E s tim a te d N o s a tis fa c to r y b re e d in g e s tim a te. G uesses p a irs ra n g e L a k e W a ira ra p a : no sa tisfa c t o r y e s tim a te. Guesses range V e r n o n L a g o o n : h ig h ly variable in 1979 T o ta l p o p u la tio n Ja nu ary P r in c ip a l b re e d in g area O K A R I T O / W E S T C O A S T O k a r it o L a g o o n (e s tu a rin e ) L A K E E L L E S M E R E L a k e E lle s m e re fre s h /b ra c k is h O T A G O / S O U T H L A N D W a itu n a L a g o o n (e s tu a rin e ) ; L a k e W a ih o la /W a ip o r i ( fre s h w a te r) ; T a ie r i R iv e r w e tla n d (fre s h w a te r) A S H B U R T O N L A K E S L a k e E m m a ; L a k e H e ro n ; b o t h fre s h w a te r M a in lo c a tio n o f n o n -b re e d e rs in b re e d in g season O k a r it o L a g o o n b u t s tr a y in g to F a re w e ll S p it n ea r e n d o f season L a k e E lle s m e re ; L a k e W a in o n o (e s tu a rin e ) ; W a s h d y k e L a g o o n (e s tu a rin e ) A w a r u a B ay ( m a r in e ) ; In v e r c a r g ill E s tu a r y a nd O ta g o H e ads (m a r in e ). S c a tte re d o v e r v a rio u s fre s h w a te r la k e s in O ta g o L a k e E m m a ; L a k e H e ro n CJI B reeding season S ta r tin g d a te v a ria b le a n d season p ro lo n g e d d u e t o f lo o d in g a nd fr e q u e n t c lo s in g o f la g o o n f r o m sea J u ly t o F e b r u a ry S e p te m b e r to N o v e m b e r. O n s e t c lo s e ly f o llo w s w a te r le v e l d e c lin e f r o m w in t e r p e a k. T h is is a r t if ic ia lly c o n tr o lle d W a itu n a : h ig h ly v a ria b le d u e to fr e q u e n t b lo c k in g o f la g o o n o u t le t to sea. S e p te m b e r to M arch. Elsew here regula r ly S e p te m b e r to N o v e m b e r S e p te m b e r t o N o v e m b e r re g u la r N e s tin g m e th o d C o lo n ia l S o lit a r y b u t s o m e tim e s c o lo n ia l a t W a itu n a S o lit a r y a nd c o lo n ia l C lu tc h sizes 4.9 in 2 3 c o lo n ia l n e s ts in : 5.0, n ests : 4.9, nests : 4.0, nests : 4.3, nests : 4.0, nests N o re c o rd s b u t average size o f 4 2 n e w ly -h a tc h e d b r o o d s in 1971 w a s 4.3 N o re c o rd s B r o o d re a rin g F a m ily b ro o d s b u t creches rep orte d 3 2 % c y g n e ts re a re d in f a m ily b r o o d s, re s t in c re c h e s a t L a k e E lle s m e re F a m ily b r o o d s b u t c re c h e s o c c a s io n a lly re p o rte d a t W a itu n a F a m ily b ro o d s C y g n e t : p r o d u c t io n ra n ge o v e r past fiv e years E s tim a te d b re e d in g p a ir s : ra n g e o v e r p a s t s ix ye a rs V a r ia b le b u t a p p a r e n tly d e c lin in g : : : : :5 0. W a itu na variable, over past five years o v e r p a s t fiv e years. T a ie r i w e tla n d s v a ria b le O ta g o lakes c o n s ta n t W a itu n a : v a ria b le, O tag o lakes: T a ie r i w e tla n d s : a n d in c re a s in g T o t a l p o p u la tio n J a n u a r y

12 Creching has been reported or observed at Lake Wairarapa, Vernon Lagoon and O karito, all areas where nesting occurs in colonies and where food is patchily distrib uted. A t Lake Whangape, food is available all over the lake and the progeny o f colonial nesting there are aji apparently reared in fam ily broods. Creching, like the colonial nesting habit, appears to be an adaptation to ensure the exploitation o f abundant but patchily distributed food in areas o f unstable water levels. Cygnet production A t Lake Ellesmere in 1976, 65% o f the cygnets which remained in fam ily broods fledged and, overall, 52.2% o f the nesting colony survived to independence. A t O karito Lagoon in 1975, 42% of the cygnets hatched were still alive when halfgrown. A t Lake Wairarapa in 1971, the average size of broods when cygnets were almost fledged was 71% o f that recorded when cygnets were newly hatched and at Lake Whangape in 1974, 72% but these are over-estimates o f survival because they exclude the possibility that some broods may have become extinct. These data are all from areas where colonial nesting occurs. A t Pukepuke Lagoon in the Manawatu district, 29 broods reared by solitary nesting swans over the years 1974 to 1978 have been follow ed. A ll cygnets hatched were reared to fledging in 20 of these broods and, overall, 87.5% of the cygnets fledged (T A Caithness pers comm). Aerial counts o f cygnets on all major breeding areas are made annually to provide an 'order of magnitude' of cygnet production. These counts are made in January when most cygnets have fledged. Estimates o f annual productions are listed in Table 2; in 1978,- the national production was approxim ately 5000, almost half o f which was raised at Lake Wairarapa, and represents a production slightly in excess o f one cygnet per breeding pair. Estimates o f survival Survival rates have been calculated only fo r the Lake Ellesmere and Lake Whangape populations (Williams 1973, 1979 and unpubl). A t Lake Ellesmere over the period 1956 to 1974, the average annual survival o f adults (five to ten year olds) was approxim ately 84%; fo r swans in their first tw o years o f life 67%; and, during years tw o to four, 78%. Bands have been returned from birds exceeding 20 years o f age. Data from Lake Whangape contrast w ith those above. No recovery o f a swan older than nine years has been made since banding commenced there in Survival during the first year was approxim ately 30%, w ith an average o f 40% over years tw o to five. These data are derived from the banding o f cygnets and have been analysed by the method o f Seber (1971). However, assumptions im p licit in the Seber model cannot be satisfied by the data and the observed band recoveries depart significantly from 158

13 model predictions. Recently derived life-table methodology (Brownie et al 1978) allows many o f the problems associated w ith exploited species (eg annual variation in band reporting) to be overcome but it demands that data derived from the banding o f cygnets be analysed in association w ith similar data obtained from the banding o f adults. A ll but one o f the swan banding programmes in New Zealand have concentrated on cygnets. Although the above estimates o f survival must be treated w ith suspicion, it is clear that real differences in survival exist between the Lake Ellesmere and Lake Whangape populations and perhaps it is not an unreasonable w orking hypothesis that survival rates d iffe r considerably between the various regional populations. Some questions Knowledge o f the biology o f C. atratus in New Zealand has many fundamental gaps. The distribution, movements and abundance o f the species throughout the country is now adequately understood but breeding biology is very poorly known. The age o f first breeding has not yet been established in the field and u ntil this is done we cannot correctly interpret the composition o f the large non-breeding population. That only about 20% breed in any year implies either that the onset of breeding is long delayed (yet in Australia some males have been sexually active in their second and th ird years Braithwaite and Frith 1969) or that not all mature adults breed every year, or both. There is also the possibility that the age of sexual m aturity varies between populations. A t some breeding areas, eg Lake Ellesmere, there are large annual variations in the number o f nests constructed (900, 1300, 2200, 1600, 750, 600 in 1974 to 1979 inclusive) yet at other areas, eg O karito Lagoon, the numbers o f nests located have varied by less than 15% over the past four years. What are the factors which determine whether a pair w ill attem pt breeding? There are almost no data on clutch sizes and whether they vary between years and nesting areas and if so, why? There are intriguing behavioural questions raised by a species which shows tw o distinct types o f social structure on nesting grounds. Why do some pairs nest solitarily and what is missing from the behavioural repertoire o f birds nesting in colonies? Can individuals adopt either nesting method at w ill or, once they first attem pt nesting, are they restricted throughout their life to that particular breeding strategy? C. atratus has a long history of exploitation as a game-bird in New Zealand. Despite this, the dynamics o f the hunting populations have never been satisfactorily studied indeed, there has been no long-term population study of C. atratus at all and there is a strong need fo r a better understanding of the ways by which hunting affects the age structure, breeding output and survival o f swans. 159

14 Extensive wetland drainage and m odification is taking place. The levels o f many lakes are being lowered or controlled and the ephemeral wetlands typ ica lly associated w ith lakes and rivers have almost all disappeared. The effect o f these changes on a species more dependent on the wetland than most o f New Zealand's w aterfow l is poorly understood because basic inform ation on such topics as habitat, food, nesting and cygnet-rearing requirements is lacking. The recent decline o f the Lake Ellesmere population (Williams 1979) and the increased tro phic state o f some Waikato and other lakes suggest that these topics need attention just as urgently as the dynamics o f C. atratus populations. Acknowledgements The presentation o f this paper at the Second International Swan Symposium was generously sponsored by A ir New Zealand Ltd and Mr T R Giles o f Consolidated Traders Ltd. Summary (Mew Zealand's Cygnus atratus inhabit the large lowland lakes, coastal lagoons, estuaries and some marine areas o f both North and South Island and Chatham Island. Swans present at marine areas, where no breeding occurs, are mostly pre-breeders; 76% at Farewell Spit were three years old or less. Of those swans present at the principal freshwater nesting areas during the breeding season, less than one-third has been recorded nesting. However, u n til the age o f firs t breeding is established it is not possible to determine whether some swans o f breeding age fail to nest each year. Studies o f dispersal using collar resightings and band recoveries suggest that swans in New Zealand may be divided into a series of regional populations between which only lim ited intermingling occurs and, as yet, no recorded interbreeding. The tim ing and method o f nesting (solitary or colonial) and length o f breeding season vary between these populations, depending m ainly on water level stability, and the methods of brood rearing are dependent on the distrib u tio n and abundance o f food in the lakes. Breeding rate and adult survival probably d iffe r considerably between populations although substantive data are lacking at present. References Braithwaite, L W and H J Frith (1969). W aterfowl in an inland swamp in New South Wales. CSIRO W ildl Res 14: Brownie, C, D R Anderson, K P Burnham and D S Robson (1978). Statistical inference from band recovery data a handbook. US Fish & W ildlife Service Resource Publications No 131. Kear, J (1972). Reproduction and fam ily life, pp in P Scott (ed) The swans. London: Michael Joseph. K irk, T (1895). The displacement of species in New Zealand. Trans N Z Inst 28: Lindsay, C J, W J Phillipps and W A Waters (1959). Birds o f Chatham Island and Pitt Island. Notornis 8: Miers, K H and M Williams (1969). Nesting o f the black swan at Lake Ellesmere, New Zealand. W ildfow l 20: Richards, E C (1950). Diary o f E C Chudleigh , Chatham Islands. Christchurch: Simpson & Williams. Seber, G A F (1971). Estimating age-specific survival rates from bird-band returns when the reporting rate is constant. Biom etrika 58: Thom pson, G M (1922). Naturalisation o f animals and plants in New Zealand. Cambridge Univ Press. 160

15 Thompson, G M (1926). W ildlife in New Zealand, Part II. W ellington: Gov't Printer. W illiams, M (1973). M ortality o f the black swan in New Zealand a progress report. W ildfow l 2 4 : W illiams, M (1977). Locations o f recoveries o f black swans, Cygnus atratus, Latham, banded at Lake Whangape and Lake Ellesmere, New Zealand. A ust W ildl Res 4: W illiams, M (1979). The status and management o f black swans, Cygnus atratus, Latham, at Lake Ellesmere since the 'W ahine'storm, A pril N Z Jl Eco! 2: M W IL L IA M S W ildlife Service Department o f Internal A ffairs W ellington New Zealand POPULATION STRUCTURE AND PRO DUCTIVITY OF. CYGNUS COLUMBIANUS COLUMBIANUS ON THE YUKON DELTA, ALASKA C P DAU Introduction The broad, fla t delta between the Y ukon and Kuskokwim Rivers covers an area of approxim ately km 2 in southwestern Alaska. Here Cygnus columbianus columbianus is an abundant nesting species occurring in higher densities than elsewhere in its breeding range (King 1973). The swans nest throughout the Yukon- Kuskokwim Delta, which extends approxim ately 376 km north to south at the coast and 320 km east to west. The m id-portion of the delta lies at about 61 30'N. Variations in population characteristics, have been demonstrated by standardized survey methods, exploratory searches and on-site nesting studies along the coastal fringe (King 1973; Clarence Rhode National W ildlife Range unpubl data). These evaluations have shown that areas w ith in 32 km o f the coastline usually support the greatest nesting densities. Lensink (1973) analysed the population structure and productivity o f C. c. columbianus on the delta between 1963 and Since 1971 several studies have expanded knowledge o f nesting waterfowl on the Y ukon-kuskokwim Delta (Boise 1977; Dau 1974; Dau and Mickelson 1979; Eisenhauer and K irkpatrick 1977; Eisenhauer 1977; Mickelson 1975; Scott 1977). These and continuing field investigations performed by the refuge staff have provided data expanding knowledge of the nesting biology o f C. c. columbianus in selected habitats along the coastal fringe (Clarence Rhode National W ildlife Range unpubl data). 161