Corella, 2004, 28(2): 43-54

SEASONAL AND DIU RNAL PATTERNS IN ABUND ANCE OF WATERBIR DS

AT A WASTE STABILIZATION POND, VICTO RIA

ANDREW J. H AMILTON12, IAIN R. TAYLOR1 and PAMELA ROGERS3

1 Applied Ornithology Group. Johnstone Centre. School of Environmental and Information Sciences, Charles Stun University, PO Box 789, Albury, New South Wales 2640

'Cuncnt address: Primary Industries Research Victoria - Knox field, Private Bag 15. Fem tree Gully Delivery Centre, Victoria 3 I 56 'Conespondingaulhor: P. Rogers. Primary Industries Research Victoria - Knoxficld. Private Bag 15, Fernlrce Gully Delivery Centre, Victoria 3156

Received: JO August 2003

Seasonal and diurnal patterns in abundance of waterbirds (excluding waterfowl: Anatidae) on a waste­stabilization pond (Pond Nine) within the Lake Borrie System at the Western Treatment Pla:-it, Werribee, Victoria were determined. The Western Treatment Plant is an important Ramsar site, mainly for the waterbird populations that occur there. Pond Nine often supports large numbers of waterbirds. However, seasonal and diurnal use of the site has received little attention to date.

The pond was used as a non-breeding site by large numbers of Straw-necked Ibises and also by more variable numbers of Australian White Ibises, Pied Cormorants, Little Pied Cormorants, Little Black Cormorants, Australian Pelicans, Silver Gulls, Eurasian Coots and Hoary-headed Grebes.

Straw-necked Ibises, Australian White Ibises, Pied Cormorants, Little Pied Cormorants, Little Black Cormorants, Silver Gulls and Australian Pelicans demonstrated crepuscular peaks of abundance. None of these species feed regularly at Pond Nine. In contrast, species that fed at or near Pond Nine, such as Eurasian Coot, Hoary-headed Grebe and consistent numbers of Black-winged Stilts were observed throughout the day.

INT RODUCTION The migration patterns of waterbirds are often studied by

monitoring the movements of individually marked birds. This method has the advantage of providing direct evidence of movements but is often limited by small sample sizes. Changes in the abundance of populations at particular sites can also be used to determine seasonal movements and, whilst such data do not provide direct evidence of movements, they can be used in conjunction with banding data to develop a more complete picture.

This paper presents information on seasonal abundance patterns of waterbirds at the Western Treatment Plant (WTP) on the Victorian coast. It is only concerned with waterbird species that do not belong to the family Anatidae (waterfowl). Waterfowl were considered separately by Hamilton et al. 2002. The data are used to make inferences on seasonal movements and also to infer the significance of the WTP in the annual life cycles of the birds. The WTP is known to be an important site for waterbirds. It is of regional or state significance for several species, including the Hoary-headed Grebe Poliocephalus po/iocephalus(state), Straw-necked Ibis Threskiornis spinicollis (state), Lillie Black Cormorant Plwlacrocorax sulcirostris (regional and possibly state) and Pied Cormorant Phalacrocorax

varius (state) (Lane and Peake 1990). There is little information on seasonal changes in abundance for these species at the WTP. Elliget (1980) presented data for a few species from March to October in 1980 and Steele ( J 996) summarized data for Eurasian Coots Fulica atra and Hoary-headed Grebes.

Daily movements to and from roosts have not been well studied for many Australian waterbirds. Marchant and Higgins (1990) concluded that no systematic information

on times of arrival and departure from roosts exists for the Australian Pelican Peleca11us conspicillatus, Great Cormorant Plwlacrocorax carbo, Little Black Cormorant, or Little Pied Cormorant Phalacrocorax 111ela110/eucos.Here we describe diurnal changes in abundance of waterbirds at Pond Nine.

MATERIALS AND METHODS

Study site

The Western Treatment Plant occupies an area of IO 85 I hectares and is situated 35 k ilometres south-west of Melbourne on the shores of Port Phillip 13ay (38°00'S. l 44°34'E) (Fig. 1 ). Observations for the present study were made at a waste stabilization pond known as Pond Nine. part of a series of ponds that make up the Lake Borrie system. Pond Nine, which covers 109 hectares, is the largest pond. It is also the only pond in the system with a stand of dead trees (mostly Melaleuca lwU'eolara) that can be used as roost sites. The average waler depth is 60 centimetres (Cartwright, unpubl. data, 1996).

Sampling protocol Sampling was undertaken on 47 dates: three times a month from 11

July 1998 to 20 June 1999 and then at approximately monthly intervals from I August 1999 to 9 August 2000. Sampling was conducted at five evenly spaced times of day: sunrise, mid-morning, midday. mid­afternoon and sunset. Typically, it took about one hour to sample the entire pond, hence sampling for each time of day started 30 minutes before the midpoint of the sampling period (e.g. 30 mins before sunrise).

Surveys were made from defined observation points on the embankment using a Lcica® Televid 77 telescope (20-60 x zoom magnification): The

_ pond was too large to be sampled from one point

and 1l was d1v1ded rnto five sections.

_Large numbers of Straw-necked Ibis and occasionally Australian WhiteIbis Threskiornfa 1110/uc,·a in the trees, and Hoary-headed Grebe on the water, were estimated groups of t 00 or I 000, depending on !lock size.

Hoary-headed Grebe was the only species that regularly dived at Pond Nine (Eurasian Coot. generally considered a partial diving species. was rarely seen diving at Pond Nine). For 1his species a preliminary s1udy

44 A. J. Hamilton, /. R. Taylor and P. Rogers: Seasonal and Diurnal Patterns in Abundance of Waterbirds Corella 28(2)

-37°

45'

-38°

00'

-38°

15'

-38°

30'

Western Treatment Plant

km

1"-1 iMP'I �

0 5 10 t

Port Phillip Bay

0 Melbourne CBD

144°

30' 144°

40' 145°

00' 145°

15'

Figure I. Map showing the lorntio11 of the Westem Treatment Pia/II.

was undertaken to determine the maximum dive kngth. The maximumdive length was found to be 20 seconds (mean = I 2.5 s. s.e. = 0.75 s). and hence in the surveys each field of view was maintained for at least20 seconds to enable any birds below the surface enough time to re­surface and be counted. Barlow ( 1976) and Best ( 1976) identifiedsimilar dive lengths for this species. Nigltr-rime observations

Night vision equipment - Fujinon® PS-9 IO Starscope ( I xmagnification) - was only available for one night during the study (25

October I 999). Observations were made at three times: ·evening·(midway between sunset and midnight), 'midnight' (midway betweensunrise and sunset). and 'early morning' (midway between midnight andsunrise). Even with the aid of night-vision apparatus. most species were too small and too far away to be identified clearly: hence observations are only presented for the two species that could clearly be identified. Australian Pelican and Pied Cormorant.

Data A11a/_vsi.1· Several species showed strong diurnal trends in abundance and for

this reason both daily maxima and mean values were plotted when

June, 2004 A. J. Hamilton, /. R. Taylor and P Rogers: Seasonal and Diurnal Patterns in Abundance of Waterbirds 45

representing seasonal abundance trends. l)aily maximum values areparticularly meaningful for species that show diurnal trends in abundance since a site could still have been important to a species at a particular time of year even if it was only used at certain times of day. Seasonal abundance data were only plotted for species where at least five individuals (for any time of day) were observed on a minimum of ten dates throughout the study.

The abundance and frequency of occurrence of species variedsubstantially. from those species often seen in the thousands (e.g. Straw­necked Ibis) to species where only one bird was recorded throughout the entire study (i.e. Great Crested Grebe Podi,·eps cris/a//1.1·). Only the more common speciL:s satisfying an arbitrary criterion of al least 50 individuals for the entire day on a minimum of five dates were includedin the analysis. The effect of time of day on abundance was analysedusing analysis of variance with date being modelled as a blocking factor. A complete census was made of each section. so data could bepooled across sections. Data were log,0 (x + I) transformed for eachspecies to improve homosccdasticity: the transformation also improvednormality of data for each species. All an;tlyses were performed using GcnStat for Windows. 5th Edition (Lawes Agricullural Trust, IACR­Rothamsted). All comparisons between means were made at the I' = 0.05 level.

RESULTS

Seasonal abundance

Straw-necked Ibis abundance peaked in mid-autumn in both 1999 and 2000, although relatively high numbers were also observed in the 1998 winter (Fig. 2). Australian White Ibis abundance followed a roughly similar trend to that of the Straw-necked Ibis, although there was no winter peak in I 998 (Fig. 3). Pied Cormorant numbers were generally highest in spring each year (Fig. 4), but patterns for all the other cormorant species were far more variable. Little Pied Cormorant numbers (Fig. 5) peaked on a few dates from late summer to late autumn in 1999 and on one date in

(/)

� .0

0 ..__(l)

.0 E ::::i

C

8000

7000

6000

a) a)

q, 0)

:'i 6_ ...., (I) (/)

a) 0) 0) 0) 0) 0) q, 0) > c � >,0 <1l <1l z ...., � �

0)

q, :'i ....,

autumn 2000. Little Black Cormorant and Australian Pelican numbers were erratic, with no obvious seasonal patterns (Figs 6, 7). Numbers of Silver Gulls Larus 11ovaeho/la11diae and Purple Swamphens Porphyrioporphyrio tended to increase in autumn in both years (Figs 8, 11). Eurasian Coot abundance peaked in autumn 2000 but numbers were relatively consistent throughout 1999, following low levels from July through to December 1998 (Fig. 9). Hoary-headed Grebe numbers showed no consistent seasonal trends; peaks in the late winter 1998 and autumn 1999 were not repeated the following year (Fig. I 0).

Diurnal abundance

There was a significant effect of time of day on abundance (P < 0.05) for all species except Eurasian Coot.

Both ibis species, all cormorant species, Australian Pelican and Silver Gull all demonstrated distinct peaks in abundance for the sunrise and sunset samples (Figs 12a, b, c, d, e, f, g). The peak abundance at sunset occurred for both Straw-necked Ibis and Silver Gull by the arrival of large flocks of birds from about 20 minutes before sunset up until sunset. Straw-necked Ibises mostly arrived from the north, from farmland, and the Silver Gulls from the south-east, from the coast. Straw-necked Ibis flocks were sometimes in the order of 5 000-7 000 birds. In contrast, Pied Cormorant and Australian Pelican numbers tended to build up gradually throughout the afternoon. For both species the mid-afternoon abundance was about 30 per cent of the sunset level (Figs 12c, f), compared to only 0.4 per cent and 2.3 per cent for Straw-necked Ibis and Silver Gull respectively. Little Pied Cormorant, Little Black

0) 0) 0 0 0 0

0) 0) 0 0 9 9 6_ > c � >, :'i (I) 0 <1l <1l (/) z ...., � � ....,

date

Figure _2. Seasonal ab1111da11ce ,,( -;'�aw-necked Ibis a'. Pond Ni11e. Dark s{iadi11g repre.\'en/.1· daily 111ea11 ab1111da11ce, s1riped shadi11g represmts daily 111{1.X111111111 ab11111/a11ce. Label., 011 x ,a,s denote ihe J,rs/ day of /he 1110111/i. The first sample date is in Ji111· wi1h 1he 'y · axis a11d 1he las/ is ,11 1he end of the ·x · mis.

46 A. J. Hamilton, I. R. Taylor and P. Rogers: Seasonal and Diurnal Patterns in Abundance of Waterbirds Corella 28(2)

1200

1000

Cf) 800 � :.0 0 ,._

:, C 400

200

0 co co co 0, 0, 0, 0, 0, 0, 0 0 0 0 � 0, 0, 0, 0, 0, � 0, 0, 0 9 9 9

:i 6. > C: � >- :i 6. > C: ni >, Q) 0 <1l Q) 0 <1l -, z <1l 2 -, z 2 -, 2 (/) -, 2

date

Figure 3. Sca.rnnal abw1da11ce of A11stralia11 White Ibis at Po11d Nine. Dark shading represe/11.\" daily 111ea11 ab1111da11ce, s1riped shading represents daily 111a.xi111w11 ab1mda11rt'. Labels 011 'x ' (Lris denote the first day of 1he 111011th. The Jirst sample date is in line wilh !he ·y ' (Uis a11d !he lase is a/ the e11d ,!( che 'x · (Lris.

..Cl

0

Cl) ..Cl E :, C

600

1 00

0 co � :i -,

co 0,

6.. Q)

co 0, 0, 0, 0, � 0, 0,

> C � >-0 <1l C1l z -, :;, 2

0, 0, 0, 0 0 0 0 � 0, 0, 0 9 9 9 ::, 6.. > c qi >, :i -, Q) 0 <1l

Cf) z -, 2 2 -,

date

Figure 4 Sea.11)11a/ ab1111da11ce oj Pied Cor11101ant at !'011d Nine Dark 1had111g 1epresen1.1 daily 111ean abwufa11re, 11nped .,had111g reproenl.1 daily ma.1111111111 ab11ndanre. Labels 011 ' .t' /L\/1 de1101e che Jir.1t day of the 1110111h The }int .\(/111ple dale 1.1 m /me w11h the 'y' 1Lr1.1 and the /a.11 ts at che end of the ·x·axis.

June, 2004 A. J. Hamilton, /. R. Taylor and P. Rogers: Seasonal and Diurnal Patterns in Abundance of Waterbirds 47

1 40

1 20

1 00 (/)

"E ..0 80 0

0) ..0 60 E ::i C

40

20

0 co

:i co co 0) 0) 0) 0) 0) 0) 0 0 0 0 0) 0) 0) 0) 0) � 0) 0) 0 0 c;, 9 a. > c � >- :5 a. > c � >- :i QJ 0 <1l <1l --, QJ 0 <1l Cll --,

(/) z --, ::,;; ::,;; (/) z --, ::,;; ::,;;

date

Figure 5. Seaso11al abundance of Lillie Pied Cor111ora111 at f'ond Nin,,. Dark shading represents daily mean ab11ndan1·e. striped slwding represellls daily 111axi11111111 11b1111rlmwe. Labels 011 •x' ,nis denote the Jirst day t!{ the 111011th. 711e .first sa111ple date is in line with the 'y' tnis and the last is at the rnd <if the 'x' axis.

700

600

500

.0 400 0

0) 300

E ::i C

co co co 0) O> 0) O> O> O> 0 0 0 0 � 0) O> O> O>

> � � O> O> 0 0 c;, 9 :i QJ 0 c � >- :5 a. > c � >- :i --, (/) z Cll ::,;; <1l --, QJ 0 <1l ::,;; (/) z --, ::,;; ::,;; --,

date

l'igurc 6. Sea1onal ab1111da11t·e of Lu tie Blad Cm mora/11 at P.t!Jl'i N111v !) '· / J J · 1 • , •

•-• , "'" 1 we 111g repffse111s , m y mean abundance, striped shading represents dail_v 111axi11111111 ab1mdance. Label.\ 011 .1 m11 denote the }1111 dav of the 11w111h The .first sample date is i11 line with the ·v · axis and the /a,· t is at 1/w rnd 1,f the ·x · axis.

48 A. J. Hamilton, I. R. Taylor and P Rogers: Seasonal and Diurnal Patterns in Abundance of Waterbirds Corella 28(2)

160

140

1 20

.... 1 00 :.0 0.... 80

E ::i 60 C

40

20

0 a) a) a) (]) (]) (]) (]) (]) (]) 0 0 0 0 q, (]) (]) (]) q, (]) q, (]) (]) 0 9 0 9 'S Cl. > C: � >- 'S Cl. > C: � >- 'S a, 0 <ll <ll a, 0 ro ro --,

(/) z -, � :;; --, (/) z -, � � -,

date

Figure 7. Sa,.rnnal ab11ndawe of A11srralic111 Pelirnn ar Pond Nine. Dark shading represents daily mean ab11ndance, striped shading represents daily maximum abundann�. LJ1bels on ·x· tuis dnwte the firs t day <�( 1he month Thf! Jirst sampli:! date is in line with the 'y' axis and the last is at the end of rh,· ·x· 1nis.

4500

4000

3500

1:: 3000

:.0 0 2500.... (!) 2000 ..0 E ::i 1500C

1000

500

0 a) a)

q, (]) (])

:5 6. > --, a, 0 (/) z

(]) (]) (]) (]) (]) (]) 0 0 0 q, (]) q, (]) (]) 0 9 c;, c;, C: � >- 'S 6. > C: <U >, :5 <ll <ll a, 0 ro � � -,

(/) z -, � � -,

date

Figure 8. Seasonal ab,mda11ce of Sliver G111/ al Pond Nine Dark shading represell/.\' daily mean abundance. s1riped shading represenrs daily 111axi11111111 abw1d11nl'e. /,nhe/s on 'x ' a.xis deno1e rhe fir.,·r day of rhe 111onrh. The jirsr .rn111ple dare is in line wirh rhe 'y' axis and the las/ is ar rhe end of rlw ·x·a xi.\.

June, 2004 A. J. Hamilton, I. R. Taylor and p Rogers: Seasonal and Diurnal Patterns in Abundance of Waterbirds 49

350

300

250 Cl)

"2 .0 200

,.__ Q) .0 1 50 E :::J

C 100

50

0 00 00 00 a, ()) ()) ()) a, ()) 0 0 0 0 q, a, ()) q, q, ()) q, a, ()) 9 9 9 9

6. > ::,.. 6. > C ro >- 'S 'S C ro 'S ro Q) 0 ro ro -, Q) 0 ro ::':! -, -,

(/) z -, ::':! ::':! (/) z -, ::':!

date

Figure 9. Seasonal abundance of Eurmian Coot at Pond Ni,w Dark shading represents da,lv 111ew1 abundance. srriped shading represems dailY nuuimwn �ibundance. Labels m, ·x· (Lris denore rhe first dar /J{ the month. T/1<' first sample dar,, is in line 1dth rhe ·.v · axis and the lasr is al the end <f rhe '.t' ruis

3500

l 3000

1 C/l 2500 "2 � 2000 0 ,.__ Q)

.0 1 500 E :::J

C

1 000

0 00

� :, -,

00 00 a, ()) a, a, a, q, 6. > c ro Q) 0 ro (/) z -, ::':!

a, a, a, a, 0 0 0 0 a, qi a, a, 0 9 <;> 9 >, 'S 6. > c ro >- 'S ro Q) 0 ro ro ::':! -,

(/) z -, ::':! ::':! -,

date

Figure I 0. Srn.rnnal ab11ndm,ce of I luarv-headed Grt'be ar Pond Ni,w Dark shading represents dail_,· 111ean ab11ndance, srriped shading represenrs dail_\' 111(Lri11111111 abw,dance. Labels on ·x · aris denme rhe .firs1 dav of tlw 11u1111h. 71,e ./ir.H .rnmph• dau' is in line ll'i th 1he ·_v · (Ltis and 1he lasr is m Iht' md ,>{ rhe ·x · 1uis.

Cormorant and Austral ian White Ibis were roughly intermediate between these extremes, with the mid-afternoon numbers being 16.7 per cent. 22.7 per cent and 12.8 per cent (respectively) of those for sunset (Figs 12d. e. b).

Of the species that showed distinct crepuscular trends in

abundance. several including both ibis species, Pied Cormorant. Australian Pelican and Si lver Gull . were significantly more numerous at sunset than they were at sunrise (Figs I 2a. b. c. f. g). This implies that many birds

left the roost before the sunrise sample . Qualitative observations of both ibis species and Australian Pelican suggested that many birds left the roost at around first light. well before the sunrise sample was taken ( i .e. > ..+5 rnins before sunrise). Moreover, during the night time study, Australian Pelican and Pied Cormorant were present in the trees at the evening sample (7 and 95 respectively) but none of either species was seen at midnight. Australian Pelican numbers had increased again to four by early­rnorning, but Pied Cormorants were still absent.

50 A. J. Hamilton, I. R. Taylor and P. Rogers: Seasonal and Diurnal Patterns in Abundance of Waterbirds Corella 28(2)

45

40

35

(/) 30 � .0

0 25

E

co co co CJ) CJ) CJ) CJ) cr, cr, 0 0 0 0

� CJ) CJ) cr, � cr, � CJ) CJ) 0 9 9 9

:5 6.. > c cu >, :5 6.. > c cu >, :5 --, Q) 0 "' "' --, Q) 0 "' "' --,

Cf) z --, � � Cf) z --, � �

date

Figure 1 1 . Seaso11a/ ab1111da11ce <!f 1'11rple Swamp/re1 1 at Pond Nine. Dark s/radi11g represe11ts daily 111ea11 ab1111da11re, striped s/radi11g represe11ts dailymc1.Xi111u111 ab1111da11ce. Labels 011 ·x· axis de11ote t/re first day of 1/re mo1 1t/r. 111<' firs/ sample date is i11 li11e wi1/r 1he 'y ' axis a11d the last is m t/re end <!f 1/ie ·x· (nis.

TABLE l List of less common waterbird species observed at Pond Nine. Maximum abundance valuesrepresent the maximum number of birds seen on any particular date at any particular time of day. Minimum refers to the minimum number of birds (other than zero) observed. For species that were observed on only one occasion. the number of birds seen is presented in the maximum columnonly. Similarly. for species that were seen on more than one occasion, but for which the samenumber of birds were seen on each of these occasions, a value is only presented in the

maximum column. The number of dates that a species was observed on is also presented. Minimum Maximum No. dates

Darter (A11/ri11ga 111e/a11ogaster) 4 I Great Crested Grebe (Podiceps rri.1w111s) I I Dusky Moorhen (Ga//i1111/a te11ebrosa) I I White-faced Heron (Egrella 11ovaelwllandiae) 2 7

Cattle Egret (Ardea ibis) 80 1 2Great Egret (Ardea alba) 3 3 Intermediate Egret (Ardea in termedia) 2 6 Australasian Bittern (Bota11rus poiciloptilris) I 2Glossy Ibis (Plegadis falci11el/us) 8 3 Royal Spoonbill (l'/atalea regia) 1 2 28Yellow-billed Spoonbill (Plata/ea J1avipes) 2 2Marsh Sandpiper (Tringa stagnatilis) I 2Sharp-tailed Sandpiper (Calidris acumi11a1a) JOO 1 5Red-necked Stint (Calidris rujico/lis) 100 6 Masked Lapwing ( Va11e//11.< miles) 1 0 4

Red-kneed Dottcrcl (Ery t/rrogo11ys ci 11c111s) 5 9 Red-necked Avocct (R�curvirostra novaehollandiae) I 250 8Pacific Gull (lams pac(fir11s) I 2 7

White-winged Black Tern (C/,/ido11ias leucopterus) 7 8 1 2

Little Tern (Sterna a/bifro11s)f'airy Tern (S1ema 11erei.,) Crested Tern (Sterna bergii)

For Hoary-headed Grebe there were significantly more birds at sunset than at any other time of day except sunrise, but there were no signi ficant differences in numbers between any other times of day (Fig. l 2i).

Numbers of Eurasian Coot were consistent throughout the day with no significant differences between any two particular times of day (Fig. I 2h) . Black-winged Stilts

I I I I

Hima11topus himantopus tended to use Pond Nine most frequently in the first half of the day (Fig. 12j).

Aside from the species mentioned thus far, several other less abundant waterbird species were observed at Pond Nine (Table 1). These 'less common' species did not satisfy the arbitrary criteria for seasonal and diurnal analysis described in the methods.

June, 2004 A. J. Hamilton, /. R. Taylor and P Rogers: Seasonal and Diurnal Patterns in Abundance of Waterbirds

Q) (.) C co

-0 C ::::,

.0 <(

3·0 ..... ( a-)-S-,--tr_a_w_-n_e_c...,.k_e_d,....l..,..b..,..

is----

3-47-

_3--,

2.5

2 .0

1 .5

1 .0

0.5

98.3

I LSDo os

1 . 5 0.6

0 · 0 +-( b_)..1.

A_u_s.Ltr_ai::I ia;;;;n;;iW---1..h-

it-e.i.,,

I b,...i

i'""s ----1. ....... __..__--t

1 .5

1 .0

0.5

2.5

2.0

1 .5

7.8

48.9

I

0.4 0.4

I

1 1 .7

1 . 5

126.9

38.6

0 1 .0 0) 2.4 2.7

0.5

o. 0 +-( d_)..1.

L-it-tleL...,,IP,-ie....,d-L:C,-0'-rm�o-ra..._n.,...t _._.....___._---1 5.0

0.75

0.50 3

.6

0.25 0.6 0.4

0.2

0 · 00 +-,.( e""')..J

L,_..,

i-tt

,.i.le--:B�l

-a""'ck,.i.....,,C'-o-rm ........ o...1ra_n...,.t ..1.-..._ ...... --1

1 .5

1 1 .0 1 0

5.6 I 2.5 0.5 1 .2 1 .4

0. 0 _.__.....,._.__._-r--'__.-r-....... ......,r-'-......... ""T"""-'----'

sr mm md ma ss

Time of Day

2.0 ...-----------::,p--,-1· --------, (f) Austral ian e 1can

1 .5

.212 1 .0

I

0.5 .22.

n 0. 0 +..--..J----....L....L--11-'-L..-................................ ___ -;(g) S i lver Gu

2.5

2.0 I 1 .5

1 .0

0.5 n4

.8

r1d, n

3

.4

rl l I 0.0 +-----'-..L.-L-..J-....L.. ............. � ....... -L..-.., (h) Eurasian Coot

1 .5

1 .0

0.5

I � 22:.!.. � � .1.!.2.

o .o +--'---_.__.___,_.d_G__....,.b ..................... _...._---t ( i) Hoary-heade re eI 399_ 9

2.5

2.0

1 .5

1 .0

,lli..6 � w �5 --

0.5

0.0 +--L-...L...L-,._,.___.___.__ ...................... _; U) Black-winged Sti lt 0.7 21.. .11. 0.6

0.5 0.4 0.3 0.2 0.1

..22. I 1 4

...L2.

0. 0 ..L..---JL...,-r--'---L-...-............................. r-'-...... -r-.__--' sr mm md ma ss

Time of Day

51

Figure 12. Diurnal thanges in ab1111da11te (logw (x + I)) of waterbirds at Po11d Nim•. sr = .rn11rise, 111111 = 111id-111omi11g, 111d = midday, ma =midaf1emoo11, .u = s1111se1. Note that the ejfert of time of day was significant for all .1pecies (P < 0.05) exrl'pt E11ra.riat1 Coot. Numbers above bars rt'presenl bark-tra11.\/on11ed 111ea11s.

52 A. J. Hamilton, I. R. Taylor and P. Rogers: Seasonal and Diurnal Patterns in Abundance of Waterbirds Corella 28(2)

DISCUSSION

Seasonal use of Pond Nine

Several species of waterbirds used Pond Nine during the apparent non-breeding season. In the species accounts below, seasonal variation in numbers is primarily discussed i n relation to possible movements associated with breeding seasons. However, it is acknowledged that other factors, such as the condition of feeding sites or erratic inland rainfall events, might also be important for determining seasonal movements.

Straw-necked Ibises breed mostly in northern or south­eastern Australia. For those breeding in the north, north­eastern Queensland appears to be a non-breeding area, while for those breeding in the south-east, south-eastern Queensland is believed to be used in the non-breeding season (Carrick 1962; McKilligan 1975; Marchant and Higgins 1990). However, it has also been suggested that some birds are sedentary around breeding sites (Marchant and Higgins 1990). In the present study, large flocks were observed at Pond Nine after the southern breeding season of August to December (McKilligan 1975) each year. This suggests that after breeding in the south-east, many birds may move to Pond Nine, a non-breeding site, rather than move north or stay around their breeding sites. The autumn flocks at Pond Nine were often large (Fig . 2) , and hence the site may be considered to be an important non-breeding resource for the species. In addition to the safe roosting habitat offered by the dead trees, the surrounding cattle pasture may add to the attractiveness of the site. Much of this pasture is flooded with sewage each summer as part of the land-filtration treatment system and is probably fertile and likely to support high i nvertebrate densities. Large flocks of Straw-necked Ibis were frequently seen feeding in these pastures.

Australian White Ibises also breed in spring and early summer (Marchant and Higgins 1990), and like the Straw­necked Ibis, they demonstrated peak abundance at Pond Nine shortly after this breeding period . Banding studies by Lowe ( 1984 - cited in Marchant and Higgins 1990) and analysis by Marchant and Higgins ( 1990) of the Australian Bird Atlas (Blakers et al. 1984) data suggest that, in contrast to the Straw-necked Ibis, this species remains in south-eastern Austral ia throughout the year. Thus, it is possible that some birds breeding in other parts of south­eastern Australia may use Pond Nine as a non-breeding s i te, although not in such high numbers as the Straw­necked Ibis. Australian White Ibises were also seen feeding in the surrounding land-filtration pastures during the day.

The Pied Cormorants reached peak abundance in spring each year. The only breeding data available for this species in south-eastern Australia comes from a three year study at Pond Nine itself i n the early 1970s (Norman 1974), when it used to breed there in high numbers. Laying started at the end of May and had finished by the start of November, and hatching was complete by the start of December. In the present study numbers i ncreased throughout winter and reached a peak around the end of spring each year. However, no nests were sighted at Pond Nine throughout the two year study.

No overall seasonal trends in abundance for Little Pied Cormorants have been reported in Victoria as a whole (Marchant and Higgins 1990), although Missen and Timms ( 1974) reported peaks in autumn and winter in western Victoria. In the present study, peak abundances were recorded from late- summer to the end of autumn in the first year (Fig. 5). In the second year, only an autumn peak was observed, although sampling frequency was less then. The timing of breeding in this species is not well understood. It has been suggested that like the Little Black Cormorant, it breeds mainly in spring and summer in the south, but it may also breed al I year round (Marchant and Higgins 1990). Therefore, it is possible that the peak in abundance in late-summer/autumn 1999, and maybe even the single peak in March 2000, represent an influx of birds after breeding.

Marchant and Higgins ( 1990) have suggested that there is a oeneral influx of Little Black Cormorants into Victoria i n s;ring and summer. This was based on an analysis of data documented in the Victorian Bird Atlas (Emison et al.I 987) . Little Black Cormorants are believed to breed in southern Australia in spring and summer, although evidence supporting this is generally anecdotal and they may apparently breed at any time of year (Marchant and Higgins 1990). There were no clear seasonal trends in abundance in the present study.

Australian Pelicans generally breed throughout the year, and according to reporting rates in the Victorian Bird Atlas (Emison et al. 1 987), there are no regular seasonal movements; breeding and movements seem to be driven by environmental factors such as rainfall (Marchant and Higgins 1 990). Abundance data from over the two years of this study (Fig. 7) tend to suggest that Pond Nine is used throughout the year as a regular roosting site. No nests were observed.

Silver Gull abundance was erratic, although there was a general tendency towards peak abundances in autumn each year, after the acknowledged spring/summer breeding season for south-eastern and eastern Australia (Marchant and Higgins 1 990). Hence, it is assumed that Pond Nine was predominantly used as a non-breeding site. Silver Gulls are generally considered to be partly sedentary and partly nomadic/migratory. There are no clear large-scale seasonal movements (Marchant and Higgins 1990). It is possible that birds arriving at Pond Nine in autumn have come from breeding grounds elsewhere in south-eastern Australia. Wheeler and Watson ( 1 963) reported that most adult and young birds departed breeding grounds at Altona, Victoria ( <50 km east of Pond Nine) in November. However, this colony is probably not used any more, and the main breeding colony is now at Mud Islands in Port Phillip Bay (P. Dann, pers. comm.; E. Walker, pers. comm.; R. Jessop, pers. comm.).

Eurasian Coot abundance peaked around February toMay each year, after the Victorian breeding season from around August to January (Bedggood 1 980). However, there was substantial variation between the first and second years of the study. Abundance during the breeding season in the second year was similar to that for the non-breeding season of the first year (Fig. 9). Other counts at Lake

June, 2004 A. J. Hamilton, /. R. Taylor and P Rogers: Seasonal and Diurnal Patterns in Abundance of Waterbirds 53

Borrie also depict a general trend of peak abundance during late summer and early spring (Steele 1996). In contrast, El l iget ( 1980) observed relatively consistent abundance throughout the duration of her March to October study at Lake Barrie. Whilst this species does not appear to undertake large-scale seasonal movements, numbers at particular sites are known to fluctuate seasonal ly (Emison e1 al. 1987; Marchant and Higgins 1990), although reasonably consistent abundances have been reported in some areas throughout the year (Leach and Hines 1987).

Marchant and Higgins ( 1990) suggest that the breeding season of Hoary-headed Grebes extends from September to January. Numbers at Pond Nine fell towards the end of the 1998 winter and stayed low from October through to the end of January, before peaking again from February through to the end of April in 1 999 (Fig. 10). This pattern possibly suggests an influx of birds arriving after breeding, but it was not repeated the following year. However. Steele ( 1996) also reported a general trend towards peak abundance at the end of summer or start of spring.

In Victoria, Purple Swamphens breed from mid-August to December (Bedggood 1970). Peak abundance was recorded in March to May each year at Pond Nine matching earlier observations by Elliget ( 1980). In New Zealand, they are not known to undertake large-scale seasonal migrations. and may stay in their territory throughout the year (Craig 1979, 1984). However. dispersal has been recorded. presumably in response to food availability (Norman and Mumford 1985). The peaks in abundance at Pond Nine after the breeding season may represent either recruitment of new birds through breeding at Pond Nine itself or an influx of birds from a breeding site elsewhere. The latter is more likely as no nests, broods or juveniles were seen at Pond Nine.

Diurnal ab1111da11ce

Straw-necked Ibises and Australian White Ibises are generally considered to arrive at roosts just before sunset, and depart around sunrise, but this has not been quantified before (Marchant and Higgins 1990). This belief is supported to a degree by the findings of the current study, where numbers of both species were clearly greatest arou�d sunrise and sunset. However, Australian White Ibises tended to arrive anytime from mid-afternoon through to sunset, whereas the large Straw-necked Ibis flocks generally arrived within the last 20 minutes before sunset.

Mi l l iner ( I 972, cited in Marchant and Higgins I 990) found that the Pied Cormorants started to arrive at the roost around 2.5 hours before sunset (roughly equivalent to 'mid­afternoon · in present study) and the majority had arrived by half an hour before sunset. The present study also supports a gradual arrival of birds over the afternoon (Fig. 12c). In contrast, Stonehouse ( 1 967), in a study of New Zealand shags, observed that most birds returned to the roost :vithin the last hour before sunset and departed within the first and second hours after sunrise. However observations from the present study suggest that many bird� may leave the roost well before sunrise.

Daily use of roosts by Little Pied Cormorants and Little Black Cormorants has not been studied in depth. Potts (1977 - cited in Marchant and Higgins 1990) made the general observation that both these species depart from the roost at around sunrise, or just after, and arrive around sunset, or just before. Data from the present study tend to support these observations, with greatest numbers of birds being reported at sunrise and sunset. Furthermore, unlike the other species that demonstrated crepuscular peaks in abundance, there was no significant difference between numbers at sunrise and sunset for these two species, which suggests that they did not general ly depart the roost before the sunrise sample.

No information exists for diurnal use of roosts by the Australian Pelican (Marchant and Higgins 1990). In the present study, it was found to have a similar diurnal abundance pattern to the cormorants, with peak numbers at Pond Nine at both ends of the day. In contrast, the Great White Pelican Peleca11L1s 011ocro1a!L1s and the Pink-backed Pelican Peleca1111s rufescens have been reported to forage most intensively around sunrise and sunset (Fasola and Canova 1993; Ntiamoa-Baidu el al. 1 998). The limited data set from the night time observations i n the present study suggests that Australian Pel icans might leave the roost, presumably to feed, during the night. This species has previously been reported to feed at night (Stone 19 I 3), although i t is generally considered to feed mostly during the day (Marchant and Higgins 1990). Night feeding also has been demonstrated for Great White Pelican (Ntiamoa­Baidu et al. 1 998) and Pink-backed Pel ican (Fasola and Canova 1993).

Of all the species demonstrating crepuscular peaks in abundance, the difference between the sunrise and sunset peaks was substantial ly greatest for the Si lver Gull . However, these data may not provide a true reflection of the diurnal movements of this species. Before sunrise, large roosting flocks were often observed on an area of land less than 40 metres from the north-eastern corner of Pond Nine. Presumably the birds that assembled on Pond Nine at sunset moved to this site later. These birds usually left this roost at around sunrise. Movement of Silver Gulls to a roost at sunset, and departure around sunrise, has been reported previously (Wheeler and Watson 1963).

Black-winged Stilts were generally observed more frequently in the morning and most of the birds observed at Pond Nine were seen foraging on the edae of the islands. However, Pond Nine did not appear to b; the main feeding site for this species; large flocks were regularly seen feeding in a shallow wetland that abuts the north­eastern corner of Pond Nine (commonly known as 'Paradise Ponds').

Co11clL1sio11

Large numbers of several species of waterbirds used pond Nine as a non-breeding site. Species that tended to feed away from Pond Nine demonstrated crepuscular peaks of abundance at Pond Nine, whereas species that used the pond as a feeding site were found there in consistent numbers throughout the day.

54 A. J. Hamilton, I. R. Taylor and P. Rogers: Seasonal and Diurnal Patterns in Abundance of Waterbirds Corella 28(2)

ACKNOWLEDGMENTS We thank Peter Dann. Ros Jessop and Eddie W,dkcr for sharing their

insights on Silver Gull breeding sites in Port Phillip flay. Weacknowledge the financial support of the Johnstone Centre. Charles StunUniversity.

REFERENCES Barlow. M. L. ( 1 976). Breeding ecology of Hoary-headed Grebe in

Southland. Nmomi., 23: 1 83-1 87.Bedggood, G. W. ( I 970). Bird notes from East Gippsland. /\11s t . Bird

Watchu 3: 252-265.Bcdggood. G. W. ( 1 980). Birdlifc between Lake Tyers and Marlo,

Victoria. A11st. Bird Watcher 8: 147-149.Best. H. A. ( 1 976). First sightings of the Hoary-headed Grebe (Podi,-ep.1·

policrpha/11s) in New Zealand. Nowrnis 23: 1 82-183.Blakcrs. M .. Davies. S. J. J. F. and Reilly. P. N. ( 1 984). 'The Atlas of

Australian Birds'. (Melbourne University Press, Melbourne.)Carrick. R. ( 1 962). Breeding. movements, and conservation of ibises

(Threskiornithidae) in Australia. CS/RO Wild/. Rn 7: 7 1-89. Cartwright. D. ( 1 996). Information (literature) review of wastewater

lagoons. with an emphasis on the Western Treatment Plant, includingLake Bo1Tic. Melbourne Water (unpublished).

Craig, J. L. ( I 979). Habitat variation in the social organisation of a communal Gallinule, the Pukcko. Behav. Ecol. Soc iobiol. 5: 331-358.

Craig. J. L. ( 1 984). Arc communal Pukcko caught in the Prisoner'sDilemma'' Behav. Ecol. Sociobio/. 14: 147-150.

Elliget. M. ( 1 980). 'A study of Lake Barrie. Werribee Sewerage Farm as a waterfowl (Anatidae) refuge area.' B. Sc. (Hons) thesis. La Trobe University, Bundoora.

Emison. W. 13., Beardsell. C. M .. Norman, F. L., Lyon, R. H. andBennell, S. C. ( 1 987). 'Atlas of Victorian Birds'. (Department of Conservation. Forests and Lands and Royal AustralasianOrnithologists Union. Melbourne.)

Fascia, M. and Canova. L. ( 1 993). Diel activity of resident andimmigrant waterbirds al Lake Turkana. Kenya. Ibis 135: 442-450.

Hamilton, A. J.. Taylor. I. R. and Hepworth. G. (2002). Activity budgets of waterfowl (Anatidac) on a waste stabilisation pond al the Western Treatment Plant, Victoria. £11111 102: 1 7 1-179.

Lane, B. and Peake, P. ( 1 990). Nature conservation al the Werribee Treatment Complex. Melbourne and Metropolitan Board of Works. Report 9 I /008 (unpublished).

Leach, G. J. and Hines. H. B. ( 1 987). Birds of the Marburg district.south-east Queensland. S1111/Jird 17: 65-77.

Marchant, S. and Higgins, P. J. (Eds) ( 1 990). 'Handbook of Australian.New Zealand and Antarctic Birds. Vol. I Ratites to Ducks.' (Oxford University Press. Melbourne.)

McKilligan. N. G. ( 1 975). Breeding and movements of the Straw-necked Ibis in Australia. £11111 75: 199-212.

Missen, R. and Timms, 13. ( 1974). Seasonal f luctuations in waterbirdpopulations on three lakes near Camperdown. Victoria. Aust. BirdWau-/,er 5: 128-1 36.

Norman, F. I. ( 1 974). Notes on the breeding of the Pied Cormorant nearWcrribce, Victoria, in I 971 . 1972, and 1 973. /:11111 74: 223-227.

Norman, F. I. and M urnford, L. ( 1985). Studies on the PurpleSwamphen. Porphyrio porphyrio, in Victoria. Aust. Wild/. Res. 12 :

263-278Ntiamoa-Baidu. Y., Picrsma, T. , Wiersma, P.. Poot, M., Banlcy, P. and

Gordon, C. ( I 998). Water depth selection, daily feeding routines anddiets of waterbirds in coastal lagoons in Ghana. Ibis 140: 89-103.

Steele, W. K. ( 1 996). An annotated bibliography of inter-relationships between waterbirds and changes in cfllucnl flows. Royal Australasian Omitho/ogisrs U11io11. Report 1 1 2.

Stone, A. C. ( 1 9 1 3). Some swamp birds. /:11111 13: 82- 86.Stonehouse, 13. ( I 967). Feeding behaviour and diving rhythms of some

New Zealand shags, Phalacrocoracidae. Ibis 109: 600-605.Wheeler, W. R. and Watson, I. ( 1963). The Silver Gull Larus

novae/wl/a11diae Stephens: Part I. Distribution and status. £11111 63:99-173.