OSTRICH 2010, 81(3): 179–188Printed in South Africa — All rights reserved

Copyright © NISC (Pty) LtdO S T R I C H

ISSN 0030–6525 EISSN 1727–947Xdoi: 10.2989/00306525.2010.517965

Ostrich is co-published by NISC (Pty) Ltd and Taylor & Francis

KwaZulu-Natal (KZN), South Africa, is the most southerly area in Africa in which the Pink-backed Pelican Pelecanus rufescens breeds. This area also boasts the most southerly breeding colony on the eastern seaboard of Africa for the Great White Pelican P. onocrotalus. Considerable amounts of data relating to the distribution and numbers of these two species in north-eastern KZN have been collected in a variety of ways by a number of people. These data were collated to estimate the distribution and status of the two pelican species and to evaluate any trends. There is no evidence of contraction or expansion in the distribution of either species. Numbers in south-eastern Africa are estimated to be 6 000–9 000 for the Great White Pelican and 600–900 for the Pink-backed Pelican. Numbers of both species were highly variable at the main locality, Lake St Lucia. For their long-term conservation it is necessary to take into account that neither population is restricted to the study area and that there are fluctuations in their distributions and numbers.

Numbers and distribution of the Great White Pelican Pelecanus onocrotalus and the Pink-backed Pelican P. rufescens in north-eastern

KwaZulu-Natal, South Africa

Meyrick B Bowker1, Richard H Taylor1,2 and Colleen T Downs1*

1 School of Biological and Conservation Sciences, University of KwaZulu-Natal, Private Bag X01, Scottsville 3209, South Africa 2 Ezemvelo KZN Wildlife, Private Bag X01, St Lucia Estuary 3936, South Africa

* Corresponding author, e-mail: downs@ukzn.ac.za

An effective conservation and management plan for any species requires certain baseline data, including an estimate of the number and distribution of the species. Here we address these two issues for the Great White Pelican Pelecanus onocrotalus (Linnaeus, 1758) and the Pink-backed Pelican P. rufescens (Gmelin, 1789), of which both are regarded as flagship species of the north-eastern regions of KwaZulu-Natal (KZN). They are large, conspic-uous, white birds, easily seen and identified and their distri-butions are limited by their having to be near sizeable water-bodies (del Hoyo et al. 1992). It is therefore possible to estimate their numbers in an area. However, they are highly mobile (Crivelli et al. 1991, Izhaki et al. 2002) and this must be considered in any estimate of numbers.

Great White Pelicans loaf and forage socially and nest on the ground in colonies (Johnsgard 1993, Crawford 2005). Because of these habits it is possible to estimate their numbers relatively accurately. Some of the loafing, foraging and breeding sites in KZN are accessible at certain times of the year and pelicans can be successfully counted from the ground or from a boat, while at other times they are inacces-sible and pelicans can only be counted from the air.

The Pink-backed Pelican is a bird of sub-Saharan Africa and the Afrotropics including southwest Arabia (Newton and Symens 1996, Ryan 2005) and the north-eastern regions of KZN. This region is home to the only regular breeding colony of these pelicans in South Africa (Taylor 2000). Even though their distribution is limited by their having to be near water-bodies, they are not as easy to census as the Great White Pelican as they are not as social (Johnsgard 1993),

can be found on smaller water-bodies (MBB pers. obs.) and are therefore more difficult to locate. They nest in the canopies of trees and are social while breeding (Burke and Brown 1970). This provides the best opportunity to estimate their numbers.

The distributions of the Great White Pelican and the Pink-backed Pelican in north-eastern KZN have been well documented on a presence or absence basis (Cyrus and Robson 1980, Williams and Borello 1997a, 1997b, Bowker and Downs 2008b). Large portions of the Great White Pelican’s and Pink-backed Pelican’s ranges lie within the area of Lake St Lucia (Williams and Borello 1997a, 1997b) and over the years many observations of these birds, including estimates of numbers, have been made and recorded by the staff of the Natal Parks Board, now Ezemvelo KZN Wildlife (EKZNW), and other interested parties. These observations were not made regularly or systematically until the advent of the Coordinated Waterbird Counts (CWAC) (Taylor et al. 1999). The records that have been made since breeding was first noted in 1933 have been used by Berutti (1980), Bowker (2006) and Bowker and Downs (2008a), but no other study has focused on these two species in this area of South Africa.

In KZN the two pelican species are fairly widespread, but much of their feeding and breeding is centred on the Lake St Lucia area. As they are long-lived birds that do not breed successfully every year in the area, it is difficult to detect changes in their numbers (Bowker 2006) and it may be many years before any change is detected. This study analysed the data available for north-eastern KZN and

Introduction

Bowker, Taylor and Downs180

reports on the numbers and the distributions of both pelican species. It was expected that their distribution range in KZN had not contracted and that they were still mainly associ-ated with Lake St Lucia.

Methods

Study area and period The study area was north-eastern KZN in South Africa, extending from the Mozambique border in the north to Durban in the south, and from the Lebombo Mountains in the west to the eastern seaboard (Figure 1). The focus was placed on the major water bodies and rivers of this region that were considered potential foraging or breeding sites for the pelicans (Figure 1). This area was chosen as it matched the distributions of the two pelican species as illustrated in Cyrus and Robson (1980) and Williams and Borello (1997a, 1997b) (Figure 2). Within this study area, the Great White Pelican nests only on islands or at sites that are naturally inaccessible to land predators. These sites are in the northern part of Lake St Lucia. Since 1989 the Pink-backed Pelican has nested only at the Nsumo Pan in the Mkhuze Game Reserve. The study extended from March 2004 to December 2005 but included data collected since 1950.

Historical dataBoth published and unpublished data from diverse sources, including historical records from files, rangers’ reports and the photographic library of EKZNW offices at St Lucia and the Mkhuze Game Reserve, were used to assess the distri-bution and numbers of both pelican species. In order to assess historical changes in the distributions of pelicans in the region, these distributions were compared to the normally accepted ranges for KZN as illustrated in authoritative field guides (Maclean 1985, Newman 1992) and bird atlases (Cyrus and Robson 1980, Williams and Borello 1997a, 1997b) to detect population expansions or contractions.

CWAC dataData from Coordinated Waterbird Counts (CWAC) were obtained from the Animal Demography Unit (ADU) at the University of Cape Town. Most of the historical data on estimates of Great White Pelican and Pink-backed Pelican numbers pertain to Lake St Lucia or the Nsumo Pan (Figure 1) in the Mkhuze Game Reserve, and most of those in turn focus on the numbers of adults and nestlings at the breeding sites (EKZNW unpublished data). The advent of the CWAC project in 1992 provided the first coordinated count for most of the areas or water bodies at which Great White Pelicans are found. These areas are counted twice a year, in summer (December–February) and in winter (June–August), and the results are collated by the ADU who also stipulate the counting methods used (Taylor et al. 1999).

KZN coordinated count In October 2004 the first concerted effort was made to obtain an accurate estimate of the number of Great White Pelicans and Pink-backed Pelicans for all the areas of their known distribution in KZN. This exercise was repeated in January and July 2005. Water levels of Lake St Lucia

were unusually low from January 2004 to December 2005. Consequently, it was not possible to count these birds from the water or the land, and aerial surveys were chosen as the most appropriate method to estimate the numbers and distri-bution of Great White Pelicans in most of the study area.

A coordinated effort was necessary simultaneously to cover all potential pelican sites in KZN. All counts were done in the morning. Some sites near Durban and Richards Bay (Figure 1) were counted from the ground by members of BirdLife Port Natal and BirdLife Zululand, respectively. More remote locations were covered from the air by aircraft belonging to members of The Bateleurs and to EKZNW. These locations included the estuaries between Durban and St Lucia, the Empangeni/Richards Bay region as far north as Lake Msingazi, pans and lakes on the Eteza and Umfolozi Rivers, the Mkhuze River and Phongolo River floodplains, Lake St Lucia and the eastern seaboard to Kosi Bay.

Monthly aerial counts Monthly aerial pelican counts from January 2004–December

N

W E

S

AFRICA

SouthAfrica

SOUTHAFRICA

KZN

Durban

RichardsBay

SWAZILAND

MOZAMBIQUE

Kosi Bay

Lake Sibaya

Sodwana BayLake Mgobezeleni

Lake BangaziNorth

Lake St Lucia

Lake BangaziSouth

NarrowsSt Lucia Estuary

Phon

golo

Flo

odpl

ain

Phongolo R

Muzi Pan

Nsumo PanMsunduzi R

Mzinene R

Hluhluwe R

Nyalazi R

Mkuze R

Lebo

mbo

Mou

ntai

ns

Enlarged areaStudy area

Figure 1: Study area in north-eastern KwaZulu-Natal indicating the entire region and detail of the northern region

Ostrich 2010, 81(3): 179–188 181

2005 were made from the EKZNW Cessna 182 Skyline towards the end of each month. The entire Lake St Lucia north of the Narrows (Figure 1) was covered from a height of about 80 m. The same aircraft, pilot and observer were used for each flight and all flights were carried out between 09:00 and 11:00. From May 2004 the Muzi and Nsumo Pans were also included in the monthly flights, mainly to include Pink-backed Pelicans there. Visibility on all flights was good, and it was possible to see large groups from considerable distances as well as spot individual birds along the flight path.

A Garmin eTrex LegendTM Personal Navigator® was synchronised with a Fujifilm FinepixTM S5000® digital camera. All groups of pelicans were photographed for subsequent counting. Synchronisation of the camera with the Garmin navigation aid (NAVAID) pinpointed the exact geograph-ical location at which each picture was taken. The aircraft’s track was recorded by the NAVAID and the digital images were downloaded to a desktop computer for analysis. Digital images were opened in Adobe PhotoDeluxe® Home Edition 4 and the number of birds was counted and recorded.

Breeding records Pelicans attract most attention when they are breeding, as they mass in impressive numbers. Most of the histor-ical data collected by EKZNW that was related to the Great White Pelican centred on the birds on Lake St Lucia when they congregated at their breeding sites. Much of this information was recorded as ‘thousands of adults’, ‘one to two thousand pairs’, ‘500 chicks’, ‘300 nesting adults’, ‘two hundred sitting adults’ and similar entries. An attempt was made to interpret these comments in a consistent way so that these data could be used to identify any changes in the breeding effort on Lake St Lucia. Breeding success for the Great White Pelican has been estimated at about 0.8–0.9 nestlings per pair taken at the time fledging occurs (Brown and Urban 1969) and in this analysis a figure of 0.85 nestlings per breeding pair per annum was used. For the Pink-backed Pelican the estimate of nest success ranges between 0.40 and 0.95 fledglings per pair (Burke and Brown 1970, Din and Eltringham 1974). A figure of 0.68 was used in the analysis and estimates of breeding effort.

When estimating the number of adults associated with ‘500 nestlings’ a figure of one nestling per breeding pair was used, as the age of the nestlings was seldom indicated and final mortality rates may not yet have been reached. Where the number of nests only was recorded, then the estimate of numbers of breeding adults has been taken as twice the number of nests.

Wetlands International (2002) considers that counting nests or the number of incubating birds is a reliable way of estimating the number of breeding pairs. One method advocated to estimate the total number of birds and to produce some level of consistency is to multiply the number of breeding pairs by three, or the number of mature adults by a factor of 1.5, as this allows for the inclusion of the immature birds in the estimates (Wetlands International 2002).

Proportion breedingNo estimates were made in this study of the fraction of breeding birds in the pelican populations. Due to the low water levels in Lake St Lucia, it was not possible to access the main loafing sites, where some estimate of the fraction of birds in breeding plumage could have been attempted.

StatisticsDescriptive statistics were calculated using StatisticaTM 7 (Statsoft, Tulsa, Oklahoma, USA).

Results

Distribution and rangeThere was no evidence from analysis of the unpublished data, from the literature and reports, or from the observa-tions during the study period, that the overall distribu-tions of the Great White Pelican and the Pink-backed Pelican in north-eastern KZN were different from those already described (Cyrus and Robson 1980, Newman 1992, Williams and Borello 1997a, 1997b). This implies that there was no contraction or expansion in their ranges. The pelicans were mainly at Lake St Lucia and its associ-ated wetlands. However, a second breeding colony of Pink-backed Pelicans has been established in trees in the Botanical Garden in Durban. This colony consists of fewer than five breeding pairs but represents a southward extension of the breeding range of this species.

Presence or absence of pelicans at foraging sitesThere was considerable variation in the presence (absence) of pelicans at different foraging sites. The CWAC monitoring reflected these variations well (Figure 3a and b). Lake St Lucia was the only site at which both pelican species were present for all counts, while at some CWAC sites in KZN, including the Kosi lakes and Lake Sibaya, pelicans were never seen.

Estimates of pelican numbers from breeding recordsEstimates were made of the number of breeding Great White Pelicans at Lake St Lucia and breeding Pink-backed Pelicans at Nsumo Pan from 1950 to 2005 for the years in which breeding was attempted. For some years, recorded

Ndumu

St Lucia

Durban

AbsentPresent

(a) (b)

Figure 2: Distribution of the (a) Great White Pelican and (b) Pink-backed Pelican in KwaZulu-Natal indicating presence or absence using data from Williams and Borello (1997b)

Bowker, Taylor and Downs182

data of the number of nests, chicks or adults produced estimates with a range of values. In each case a compro-mise estimation was made. Values for the Great White Pelican varied from 200 to 4 400 (Figure 4a), while for the Pink-backed Pelican they ranged from 14 to 420 (Figure 4b). For those years when estimates were made, i.e. when birds attempted to breed, the average number (±SE) of Great White Pelican breeders was 1 969 ± 1 128 (n = 40) and for Pink-backed Pelican 137 ± 72 (n = 35).

Summer, winter and overall estimates of pelican numbersNumbers of pelicans counted on CWAC surveys varied considerably during and between summers (December–Febuary) and winters (June–August) for both species (Figure

5a and b). The only year in which the summer count for the Great White Pelican exceeded the winter count of the same year was 1995. Estimates of total numbers during summer counts for the Great White Pelican varied from 7 in 2001 to 3 579 in 1995, with an average of 703 ± 1 118 (n = 14). Estimates of total numbers during winter ranged from 221 in 1995 to 5 081 in 1992, with an average of 1 944 ± 1 571 (n = 14). For the Pink-backed Pelican, estimates of total numbers during summer counts exceeded estimates made during winter counts in all years except 1998 and 1999. Summer estimates of total numbers for the Pink-backed Pelican varied from 34 in 1999 to 250 in 2003, with an average of 122 ± 69 (n = 14). Winter total numbers ranged from 46 in 2003 to 168 in 1998, with an average of 82 ± 40 (n = 14).

20

45

17

57

100

14 13

57 56

70

414

4

71

25

83

17

54

11

44

Collin

's La

keDu

rban

Bay

head

NHS

Hluh

luwe

Floo

dplai

nLa

ke E

teza

Lake

St L

ucia

Mav

uya

Pan

Mfa

zana

Pan

Mfo

lozi E

stuar

yNe

she

Pan

Nsum

o Pa

nnT

shan

etsh

e Pa

n

Nyam

ithi P

an (N

dum

o)Po

ngola

poor

t Dam

Rich

ards

Bay

Wet

lands

Sund

umbil

i Sew

age

Wor

ksTh

ulazih

leka

Pan

Tuge

la Ri

ver M

outh

Umge

ni Ri

ver E

stuar

y

Umvo

ti Rive

r Estu

ary

Yeng

weni

Pan

17

67

145

100

1710 6

29

48

4

38

52

20 18

62

13

65

17 17 1711 10

44

0

20

40

60

80

100

Banz

i Pan

(Ndu

mo)

Hluh

luwe

Floo

dplai

nLa

ke E

teza

Lake

Mfu

tulul

uLa

ke S

t Luc

iaLe

njane

wet

land

Mav

uya

Pan

Mbo

zam

bo W

aste

Wat

erM

foloz

i Estu

ary

Nesh

e Pa

nNh

lonhle

la Pa

n

North

ern

Trea

tmen

t Wor

ksNs

umo

Pan

nTsh

anet

she

Pan

Nyam

ithi P

an (N

dum

o)

Rich

ards

Bay

Wet

lands

Sund

umbil

i Sew

age

Wor

ksTh

ulazih

leka

Pan

Tuge

la Ri

ver M

outh

Umge

ni Ri

ver E

stuar

y

Umhla

nga

Trea

tmen

t Wor

ks

Umvo

ti Rive

r Estu

ary

Wat

erfo

rd F

arm

Dam

sYe

ngwe

ni Pa

n

CO

UN

TS P

RE

SE

NT

(%)

20

40

60

80

100

CO

UN

TS P

RE

SE

NT

(%)

(a) Great White Pelican

(b) Pink-backed Pelican

SITE

Figure 3: Percentage of CWAC counts at which a) Great White Pelican and b) Pink-backed Pelican were present at sites in north-eastern KZN for the years 1992 to 2005. Only sites counted five or more times have been included and those sites at which pelicans were never recorded have been omitted

Ostrich 2010, 81(3): 179–188 183

The average number of individuals of a species was calculated for those sites at which Great White Pelicans (Figure 6a) or Pink-backed Pelicans (Figure 6b) were seen during the CWAC periods of summer 1992 through to winter 2005. The average number of Great White Pelicans at Lake St Lucia, 817 ± 55 (n = 26), far exceeded that at all other sites. For the Pink-backed Pelican, average totals of 40 ± 4 (n = 26) at Lake St Lucia and 43 ± 4 (n = 27) at Nsumo Pan were the highest.

Lake St Lucia accounted for most of the Great White Pelicans in north-eastern KZN in most CWAC counts conducted over the thirteen-year period (Figure 7). For the summer counts, the lake accounted for an average of 71% of the Great White Pelicans recorded in KZN, while the same value in winter was 80%. CWAC counts showed that the Nsumo Pan in summer (Figure 8a) and Lake St Lucia in winter (Figure 8b) had high numbers of Pink-backed Pelicans.

KZN pelican countsThree independent counts conducted on 16 October 2004, 29 January 2005 and 23 July 2005 of all the potential Great White Pelican and Pink-backed Pelican sites in KZN produced totals of 2 362, 2 611 and 3 929 for Great White Pelican and 46, 628 and 159 for Pink-backed Pelican, respec-tively (Table 1). Average numbers of Great White Pelican and

Pink-backed Pelican for these three provincial counts were 2 967 ± 224 (n = 3) and 278 ± 436 (n = 3), respectively.

Monthly aerial surveys of Lake St Lucia, Nsumo Pan and Muzi PanMonthly aerial surveys of Lake St Lucia during 2004 and 2005 showed variation from a minimum of 123 to a maximum of 6 558 in the number of Great White Pelicans (Figure 9a). For Pink-backed Pelicans the surveys included Lake St Lucia and Nsumo and Muzi Pans, and the variation in numbers ranged from a minimum of 0 to a maximum of 622 (Figure 9b).

Discussion

Distribution Distributions recorded in this study did not indicate any contraction or extension of the distributions compared with the atlas information (Cyrus and Robson 1980, Williams and Borello 1997a, 1997b). However, it is clear that the Great White Pelican favours some areas and that not all water bodies in north-eastern KZN are used by these birds (CWAC unpublished data). As pelicans are highly mobile, it is certain that both species range into Mozambique. Parker (1999) reported that the number of Great White Pelicans in southern Mozambique may exceed 3 000 at times and that they are

50010001500200025003000350040004500

1950

1952

1954

1956

1958

1960

1962

1964

1966

1968

1970

1972

1974

1976

1978

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

YEAR

NU

MB

ER

OF

BR

EE

DIN

G P

ELI

CA

NS

NumberNo data

(a) Great White Pelicans

50

100

150

200

250

300

350

400(b) Pink-backed Pelicans

Figure 4: Estimates of the numbers of breeding (a) Great White Pelican individuals and (b) Pink-backed Pelican individuals present at nesting colonies in north-eastern KwaZulu-Natal from 1950 to 2005

Bowker, Taylor and Downs184

common visitors to the larger water-bodies of this area. He mentioned that the Great White Pelican possibly breeds further north in the Zambezi River delta. This is corroborated by reports of occasional breeding there in the late 1990s and early 2000s (R Beilfuss unpublished data). Parker (1999) described the Pink-backed Pelican as an uncommon visitor to Mozambique, usually in small numbers but possibly exceeding 400 at times, and added that more than 200 birds may have bred at Massingir in April 1995. It is not known whether these two pelican species form discrete populations in the sub-Zambezi region of south-eastern Africa.

Importance of Lake St Lucia and Nsumo PanGreat White Pelican counts at Lake St Lucia are a good proxy for estimating numbers for the whole of KZN, as the lake is the preferred site for both foraging and breeding (Figure 6a). There may be occasions when large groups are away from the lake, as has been shown with single CWAC records of 1 000 birds at each of Lake Mfutululu and Hlonhlela Pan (Taylor et al. 1999). For the Pink-backed Pelican, CWAC records show that two sites are favoured (Figure 6b), Nsumo Pan for summer breeding (Figure 8a) and the Lake St Lucia for winter foraging (Figure 8b). The three independent coordinated counts showed that Lake

St Lucia was by far the most important water-body in the region, and always accounted for the highest percentage of the numbers of both pelican species (Table 1).

Variations of estimates, and seasonalityEstimates of numbers of Great White Pelicans in north-eastern KZN have varied considerably, ranging from 0 to 6 000. The variation may result from different methods being applied (see below), variations in counting conditions and real fluctuations in numbers. In September 1990, between 4 000 and 6 000 birds were counted on the lake (EKZNW unpublished data). Previous estimates of numbers at Lake St Lucia have been of about 2 000 pairs (Crawford et al. 1981) and 3 000 breeding pairs (Bowker and Downs 2008b). For the Pink-backed Pelican, it was estimated that 250 pairs (Williams and Borello 1997a) or 240−300 birds (Taylor 2000) bred. In this study the number of Pink-backed Pelicans breeding was about 200 pairs in 2005, and it appears that the size of the colony at Nsumo Pan may be increasing (Bowker and Downs 2008a).

The three coordinated counts between October 2004 and July 2005 showed that numbers of Great White Pelicans at Lake St Lucia ranged between 2 362 and 3 929, representing 97.0% and 98.6% of the KZN estimate, respectively (Table 1). These three coordinated counts were done at a time when the number of Great White Pelicans on Lake St Lucia was recovering after the drought that ended in early 2004 (Bowker and Downs 2008b). By September 2005, Great White Pelican numbers had recovered to approximately 6 500, probably about the maximum for the lake based on all estimates made since 1950 (Bowker and Downs 2008b). Results of the monthly aerial counts of 2004 and 2005 suggested that numbers of Great White Pelican at Lake St Lucia increased steadily from the start of 2004 to September 2005, following a rise in water levels in Lake St Lucia after rains in early 2004 and the improved foraging that accompanied this change (Bowker and Downs 2008b, Figure 9a).

The three coordinated counts of Pink-backed Pelicans in 2004 and 2005 were between 46 and 628 individuals (Table 1). The minimum value can be ascribed to the dispersal of birds to more favourable foraging areas outside of north-eastern KZN, while the maximum value is higher than previously suggested but less than the 750 individuals obtained by multiplying the 250 pairs estimated by Williams and Borello (1997a) by three, as suggested by Wetlands International (2002). Monthly aerial counts in 2004 and 2005 indicated that numbers of Pink-backed Pelicans at Lake St Lucia and Nsumo and Muzi Pans increase during the summer breeding season, and that many of these birds leave these sites when not breeding (Figure 9b).

Factors contributing to variations in estimates of numbersMany pelican counts have been done in the study area since 1950. The methods used have varied enormously, and while these records give some indication of the number of pelicans in the lake system at various dates, it is difficult to make comparisons of these data. Many of the counts were restricted to Lake St Lucia and so cannot be used to estimate accurately the numbers in the entire study area. Potentially productive sites such as the Phongolo River

1071

658

169

3579

127

70 8

589

88 7 129

207 53

7

2611

5081

1924

659

221

1646

1308

476

2829

4060

1152

656 97

8

2300

3929

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

Summer count for KZNWinter count for KZN

156

144

132

102

200

70

43 34

81

223

163

250

113

628

143

69

104

72

141

61

168

94

61 68

124

46

68

159

100

200

300

400

500

600

TOTA

L N

UM

BE

R

2 000

1 000

3 000

4 000

5 000

YEAR

(a) Great White Pelicans

(b) Pink-backed Pelicans

Figure 5: Total number of (a) Great White Pelicans and (b) Pink-backed Pelicans at north-eastern KwaZulu-Natal CWAC sites during the summer and winter counts for each year from summer 1992 to winter 2005

Ostrich 2010, 81(3): 179–188 185

floodplain were omitted in most counts (Bowker and Downs 2008c). Counts were done during different months of the year, when the distribution of the birds within the study area may have changed.

Lake St Lucia is a large, shallow lake. Many areas are difficult to access and it is possible to miss birds, especially when counting from the ground or from a boat. Great White Pelicans are social and often form groups of several thousand. If a group is in an inaccessible area and is not counted, the error can be significant. Pink-backed Pelicans are solitary and difficult to find on a lake the size of Lake St Lucia and in small water-bodies scattered throughout the study area. They are also often overlooked if they are perched in trees.

Biological and ecological factors contributing to variations in estimates of numbersAs pelicans are highly mobile, they can move in and out of the study area in large numbers if conditions are not conducive to breeding or feeding (Williams and Borello 1997b). The subadults do not necessarily associate with the adults and tend to disperse to different areas (Guillet and Crowe 1981). It is assumed that the KZN subadults range north towards and into Mozambique, as far as the Zambezi River (Parker 1999, Barnes 2000). These emigrations create real fluctuations in numbers.

The number of breeding Great White Pelicans at Lake St Lucia varies markedly, and this can be related to lake conditions (Bowker and Downs 2008b). When conditions

Collin

's La

ke

Durb

an B

ayhe

ad N

HSHl

uhluw

e Fl

oodp

lain

Lake

Ete

zaLa

ke S

t Luc

iaM

avuy

a Pa

nM

faza

na P

anM

foloz

i Estu

ary

Nesh

e Pa

nNs

umo

Pan

nTsh

anet

she

Pan

Nyam

ithi P

an (N

dum

o)Po

ngola

poor

t Dam

Rich

ards

Bay

Wet

lands

Sund

umbil

i Sew

age

Wor

ksTh

ulazih

leka

Pan

Tuge

la Ri

ver M

outh

Umge

ni Ri

ver E

stuar

y

Umvo

ti Rive

r Estu

ary

Yeng

weni

Pan

10

30

800

Banz

i Pan

(Ndu

mo)

Hluh

luwe

Floo

dplai

nLa

ke E

teza

Lake

Mfu

tulul

uLa

ke S

t Luc

iaLe

njane

wet

land

Mav

uya

Pan

Mbo

zam

bo W

aste

Wat

erM

foloz

i Estu

ary

Nesh

e Pa

nNh

lonhle

la Pa

n

North

ern

Trea

tmen

t Wor

ksNs

umo

Pan

nTsh

anet

she

Pan

Nyam

ithi P

an (N

dum

o)

Rich

ards

Bay

Wet

lands

Sund

umbil

i Sew

age

Wor

ksTh

ulazih

leka

Pan

Tuge

la Ri

ver M

outh

Umge

ni Ri

ver E

stuar

y

Umhla

nga

Trea

tmen

t Wor

ks

Umvo

ti Rive

r Estu

ary

Wat

erfo

rd F

arm

Dam

sYe

ngwe

ni Pa

n

AVE

RA

GE

NU

MB

ER

20

40

600

400

200AVE

RA

GE

NU

MB

ER

(a) Great White Pelican

(b) Pink-backed Pelican

SITE

Figure 6: Average number (+SE) of (a) Great White Pelicans and (b) Pink-backed Pelicans present at CWAC sites in north-eastern KwaZulu-Natal during summer and winter counts for the period 1992 to 2005. Only sites counted five or more times have been included and those sites at which pelicans were never recorded have been omitted

Bowker, Taylor and Downs186

are ideal for breeding, the lake might attract more than 6 000 breeding birds, but in other years there may be no breeding at all (Bowker and Downs 2008a). For both species fluctuations in numbers breeding are probably

driven by the availability of food (Kingsford et al. 1999) or nest sites (Clauss 1972, Berry et al. 1973). Some counts based on the number of breeding pelicans may fail to take into account non-breeders. Numbers might increase or decrease over a period of time if the birds have several consecutive successful or unsuccessful breeding seasons (Brown and Urban 1969).

The Great White Pelican breeding sites in north-eastern KZN are in a wilderness area, are difficult to access, and the birds can easily start their breeding unnoticed. The nests of the Pink-backed Pelicans are high, difficult to monitor and often overlap one another. An added difficulty in estimating the number of breeding Pink-backed Pelicans in the region is that they breed asynchronously (Burke and Brown 1970) and their breeding season may last for more than 6 months (Berruti 1980, Bowker 2006). Indeed, even if both species are monitored regularly, it is difficult to get accurate estimates of the number of birds that nest over the entire season because of pulses of breeding activity producing multiple cohorts of young birds.

Use of CWAC dataUsing CWAC data only is not an adequate method for estimating the total number of pelicans in the region. Some water-bodies in KZN are not counted, and the totals are given in broad categories of ‘summer’ and ‘winter’. Furthermore, counts do not take place at all sites on the same day, being in some cases several weeks apart. Thus, birds could be counted more than once, or missed altogether. As expected, analyses of CWAC records show high fluctuations in estimates of numbers of Great White Pelicans and Pink-backed Pelicans in north-eastern KZN.

Considerations for the futureProtocols need to be established to standardise the counting and monitoring of the two pelican species in KZN and the overall range of these birds needs to be established. Only three Pink-backed Pelicans and four Great White Pelicans have ever been ringed in KZN and no recoveries have been recorded (ADU unpublished data). A satellite

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Summer count for KZN

Winter count for KZNSummer count for Lake St Lucia

Winter count for Lake St Lucia

TOTA

L N

UM

BE

R

2 000

1 000

3 000

4 000

5 000

YEAR

Great White Pelicans

Figure 7: Total number of Great White Pelicans counted at all north-eastern KwaZulu-Natal CWAC sites (including Lake St Lucia) compared to the total number counted at Lake St Lucia, for summer and winter each year from 1992 to 2005

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

Other KZN sitesCount for NsumoCount for Lake St Lucia

100

175

150

125

100

75

50

25

200

300

400

500

600

TOTA

L N

UM

BE

R O

F P

INK

-BA

CK

ED

PE

LIC

AN

S

YEAR

(a) Summer

(b) Winter

Figure 8: The total number of Pink-backed Pelicans counted at all north-eastern KwaZulu-Natal CWAC sites (including Lake St Lucia) compared to the total numbers counted at Lake St Lucia and the Nsumo Pan for (a) summer and (b) winter of each year from1992 to 2005

Ostrich 2010, 81(3): 179–188 187

telemetry study for both species could give some insight into their true ranges and geographical bounds and suggest whether discrete populations of the species exist in the sub-Zambezi region of south-eastern Africa. If these popula-tions are discrete, then this might allow for more accurateestimates of numbers.

Present levels of abundanceAssuming the populations are discrete and that there are no other breeding sites within their ranges, then for the Great White Pelicans an estimate of 6 000–9 000 individuals might be acceptable, while for Pink-backed Pelicans an estimate of 600–900 is suggested. The great variation in numbers for both species indicates that there is mass movement in and out the study area. Pink-backed Pelicans seem to be increasing in north-eastern KZN, whereas numbers of Great White Pelicans appear stable.

Acknowledgements — We thank the following for their input into this study: the ADU for CWAC data; EKZNW for making their data available and for assistance with the aerial surveys; Greg Nänni for his expertise and assistance in monthly aerial counts; Paul Dutton, Donovan Barton-Hobbs, Chris Rattray, James Hopcroft, and the late Nora Kreher, all of The Bateleurs, for their voluntary assistance with the three provincial pelican counts; BirdLife Zululand and BirdLife A

rea

coun

ted

Oct

ober

200

4O

ctob

er 2

004

Janu

ary

2005

Janu

ary

2005

July

200

5Ju

ly 2

005

GW

P%

PB

P%

GW

P%

PB

P%

GW

P%

PB

P%

Nor

ther

n Tr

eatm

ent W

orks

0

0.0

00.

00

0.0

00.

013

0.3

00.

0U

mge

ni R

iver

Mou

th0

0.0

00.

01

0.0

00.

00

0.0

00.

0B

otan

ical

Gar

dens

00.

04

8.7

00.

014

2.2

00.

015

9.4

Bay

Hea

d H

erita

ge S

ite0

0.0

36.

51

0.0

00.

00

0.0

127.

5Tu

gela

Est

uary

40.

20

0.0

00.

00

0.0

00.

00

0.0

Oth

er e

stua

ries

north

to S

t Luc

ia0

0.0

00.

00

0.0

00.

00

0.0

00.

0S

anct

uary

(Ric

hard

s B

ay)

451.

90

0.0

100.

45

0.8

00.

02

1.3

All

othe

r wat

er b

odie

s ar

ound

Ric

hard

s B

ay0

0.0

00.

00

0.0

00.

011

0.3

00.

0O

ther

wat

er b

odie

s on

Um

folo

zi R

iver

and

sur

roun

ds0

0.0

00.

00

0.0

00.

00

0.0

00.

0La

ke S

t Luc

ia2

291

97.0

2452

.22

599

99.5

189

30.1

3 87

598

.625

15.7

Mkh

uze

flood

plai

n6

0.3

1430

.40

0.0

00.

030

0.8

00.

0N

sum

o P

an6

0.3

12.

20

0.0

407

64.8

00.

010

6.3

Pho

ngol

o flo

odpl

ain

(incl

udin

g P

ongo

la D

am a

nd N

dum

u)10

0.4

00.

00

0.0

132.

10

0.0

9559

.7O

ther

s: S

ibay

a, K

osi,

Ban

ghaz

i, et

c.0

0.0

00.

00

0.0

00.

00

0.0

00.

0To

tals

2362

100

4610

026

1110

062

810

03

929

100

159

100

Tabl

e 1:

Cou

nt d

ata

from

the

thre

e co

ordi

nate

d G

reat

Whi

te P

elic

an a

nd P

ink-

back

ed P

elic

an c

ensu

ses

of K

waZ

ulu-

Nat

al (O

ctob

er 2

004,

Jan

uary

200

5 an

d Ju

ly 2

005)

. Loc

atio

ns o

f are

as

coun

ted

are

indi

cate

d in

Fig

ure

1. G

WP

= G

reat

Whi

te P

elic

an; P

BP

= P

ink-

back

ed P

elic

an; %

= p

erce

ntag

e of

the

tota

l num

ber o

f pel

ican

s co

unte

d in

all

site

s fo

r tha

t cen

sus

Jan.

200

4Fe

b. 2

004

Mar

. 200

4A

pr. 2

004

May

200

4Ju

n. 2

004

Jul.

2004

Aug

. 200

4S

ep. 2

004

Oct

. 200

4N

ov. 2

004

Dec

. 200

4Ja

n. 2

005

Feb.

200

5M

ar. 2

005

Apr

. 200

5M

ay 2

005

Jun.

200

5Ju

l. 20

05A

ug. 2

005

Sep

. 200

5O

ct. 2

005

Nov

. 200

5D

ec. 2

005

100

200

300

400

500

600

TOTA

L N

UM

BE

R

2 000

1 000

3 000

4 000

5 000

6 000

MONTH

(a) Great White Pelicans

(b) Pink-backed Pelicans

Figure 9: The total number of (a) Great White Pelicans on Lake St Lucia and (b) Pink-backed Pelicans on Lake St Lucia, Nsumo Pan and Muzi Pan for the period January 2004 to December 2005

Bowker, Taylor and Downs188

Port Natal for various counting exercises and support; Caroline Fox and Bheka Nxele for access to relevant information; and the National Research Foundation for funding.

References

Barnes KN. 2000. The Eskom red data book of South Africa, Lesotho and Swaziland. Johannesburg: BirdLife South Africa.

Berruti A. 1980. The status and review of waterbirds breeding at Lake St Lucia. Lammergeyer 28: 1−19.

Berry HH, Stark HP, van Vuuren AS. 1973. White Pelicans Pelecanus onocrotalus breeding on the Etosha Pan, South West Africa, during 1971. Madoqua Series 1(7): 17−31.

Bowker MB. 2006. Conservation of the Great White Pelican (Pelecanus onocrotalus) and the Pink-backed Pelican (P. rufescens) in south eastern Africa. MSc thesis, University of KwaZulu-Natal, Pietermaritzburg, South Africa.

Bowker MB, Downs CT. 2008a. Fluctuations in numbers of Great White Pelicans at Lake St Lucia in response to changing water-levels. African Journal of Ecology 46: 282−290.

Bowker MB, Downs CT. 2008b. Breeding incidence of the Great White Pelican Pelecanus onocrotalus and the Pink-backed Pelican P. rufescens in south-eastern Africa from 1933 to 2005. Ostrich 79: 23–35.

Bowker MB, Downs CT. 2008c. The status of habitat of Great White and Pink-backed Pelicans in northeastern KwaZulu-Natal, South Africa: a review. African Zoology 43: 90–98.

Brown LH, Urban EK. 1969. The breeding biology of the Great White Pelican Pelecanus onocrotalus roseus at Lake Shala, Ethiopia. Ibis 111: 199−237.

Burke VEM, Brown LH. 1970. Observations on the breeding of the Pink-backed Pelican Pelecanus rufescens. Ibis 112: 499−512.

Clauss B. 1972. Pelicans at Lake Ngami. Ostrich Supplement 43: 176.

Crawford RJM. 2005. Great White Pelican Pelecanus onocrotalus. In: Hockey PAR, Dean WRJ, Ryan PG (eds), Roberts birds of southern Africa (7th edn). Cape Town: Trustees of the John Voelcker Bird Book Fund. pp 614−615.

Crawford RJM, Cooper J, Shelton PA 1981. The breeding popula-tion of White Pelicans Pelecanus onocrotalus at Bird Rock platform in Walvis Bay, 1949–1978. Fisheries Bulletin South Africa 15: 67−70.

Crivelli AJ, Leshem Y, Mitcher T, Jerrentrup H. 1991. Where do the Palaearctic Great White Pelicans Pelecanus onocrotalus presently overwinter? Revue d’Ecologie (La Terre et la Vie) 46: 145−171.

Cyrus DP, Robson NF. 1980. Bird atlas of Natal. Pietermaritzburg: University of Natal Press.

Del Hoyo J, Elliott A, Sargatal J. 1992. Handbook of the birds of the world, vol. 1: Ostrich to ducks. Barcelona: Lynx Edicions.

Din NA, Eltringham SK. 1974. Breeding of the Pink-backed Pelican Pelecanus rufescens in Ruwenzori National Park, Uganda. Ibis 116: 477−493.

Guillet A, Crowe TM. 1981 Seasonal variation in group size and dispersion in a population of Great White Pelicans. Le Gerfaut 71: 185−194.

Izhaki I, Shmueli M, Arad Z, Steinberg Y, Crivelli A. 2002. Satellite tracking of migratory and ranging behaviour of immature Great White Pelicans. Waterbirds 25: 295−304.

Johnsgard PA. 1993. Cormorants, darters and pelicans of the world. Washington, DC: Smithsonian Institution Press.

Kingsford RT, Curtin AL, Porter J. 1999. Water flows on Cooper Creek in arid Australia determine ‘boom’ and ‘bust’ periods for waterbirds. Biological Conservation 88: 231−248.

Maclean G. 1985. Roberts’ birds of southern Africa (5th edn). Cape Town: Trustees of the John Voelker Bird Book Fund.

Newman K 1992. Newman’s birds of southern Africa. Halfway House: Southern Book Publishers.

Newton SF, Symens P. 1996. The status of the Pink-backed Pelican Pelecanus rufescens and the Great White Pelican P. onocrotalus in the Red Sea: the importance of Saudi Arabia. Colonial Waterbirds 19: 56−64.

Parker V. 1999. The atlas of birds of Sul do Save, southern Mozambique. Cape Town: Avian Demography Unit; and Johannes-burg: Endangered Wildlife Trust.

Ryan PG. 2005. Pink-backed Pelican Pelecanus rufescens. In: Hockey PAR, Dean WRJ, Ryan PG (eds), Roberts birds of southern Africa (7th edn). Cape Town: Trustees of the John Voelcker Bird Book Fund. pp 615−617.

Taylor PB, Navarro RA, Wren-Sargent M, Harrison JA, Kieswetter SL. 1999. Total CWAC report: coordinated waterbird counts in South Africa, 1992–1997. Cape Town: Avian Demography Unit.

Taylor RH. 2000. Pink-backed Pelican Pelecanus rufescens. In: Barnes KN (ed.), The Eskom red data book of birds of South Africa, Lesotho and Swaziland. Johannesburg: BirdLife South Africa. pp 62−63.

Wetlands International 2002. Waterbird population estimates (3rd edn). Wetlands International Global Series No. 12. Wageningen, Wetlands International.

Williams AJ, Borello WD. 1997a. Great White Pelican Pelecanus onocrotalus. In: Harrison JA, Allan DG, Underhill LG, Herremans M, Tree AJ, Parker V, Brown CJ (eds), The atlas of southern African birds, vol. 1: Non-passerines. Johannesburg: BirdLife South Africa. pp 24−25.

Williams AJ, Borello WD. 1997b. Pink-backed Pelican Pelecanus rufescens. In: Harrison JA, Allan DG, Underhill LG, Herremans M, Tree AJ, Parker V, Brown CJ (eds), The atlas of southern African birds, vol. 1: Non-passerines. Johannesburg: BirdLife South Africa. pp 26−27.

Received January 2010, accepted May 2010Editor: RJM Crawford