Page 264 - Atlas Sea Birds Ver1

2.2.4.16 Breeding by seabirds at Lake St Lucia, KwaZulu-Natal

Authors: C Fox, RH Taylor, and RJM Crawford
Abstract:
Lake St Lucia lies on the central Zululand coast within the iSimangaliso Wetland Park at the northeast of South Africa. Four species
of seabird have bred at 21 sites around the lake. Maximum numbers recorded breeding were c. 3,700 pairs of Great White Pelican
(Pelecanus onocrotalus) in 2012, c. 300 pairs of White-breasted Cormorant (Phalacrocorax lucidus) in 1964, c. 1,500 pairs of Grey-
headed Gull (Chroicocephalus cirrocephalus) in 1992 and c. 750 pairs of Caspian Tern (Hydroprogne caspia) in 1972.

Introduction:
Four species of seabird have bred at 21 sites around Lake St Lucia, which lies on the central Zululand coast and is protected
within the iSimangaliso Wetland Park World Heritage Site and the St Lucia Ramsar Site. They are: Great White Pelican (Pelecanus
onocrotalus), White-breasted Cormorant (Phalacrocorax lucidus), Grey-headed Gull (Chroicocephalus cirrocephalus), and Caspian Tern
(Hydroprogne caspia). Lake St Lucia is an estuarine lake almost 70 km long and 18 km at its widest point, which has a surface area of
225–417 km depending on the water level. It extends from the Ozabeni Wilderness in the north to Cape St Lucia in the south and from the
2
Indian Ocean inland to west of False Bay (Figure 1). The park encompasses all of the lake as well as substantial portions of surrounding
grassland, savanna, coastal forest, dunes and urban habitats (Johnson et al. 1998, Bowker and Downs 2012). The seabird nesting sites
comprise islands, pans, trees and marshes. Substantial fluctuations occur in the lake’s water level and influence the availability of suitable
habitat. There is also large spatial and temporal variability in salinity and food availability in response to wet-dry periods (Bowker and
Downs 2012). The four seabirds all breed at inland as well as coastal habitats and they may move between different regions to breed
(Hockey et al. 2005).
From 1975 to 2021 Ezemvelo KwaZulu-Natal (EKZN) St Lucia. For White- breasted Cormorant the maximum
Wildlife undertook 109 counts of 114 different waterbird ratio was increased to 12 to improve sample size. For
species that use Lake St Lucia. No counts were made this (and the other) species it is likely that in some years
in 1977, 1982 and 2019. In other years the numbers of not all breeding colonies were counted. The relationships
counts ranged from 1–5 (Table 1). All counts were made between maximum counts (X) and numbers of breeding
on a single day. The lake was divided into sections and pairs (Y) are shown in Figure 2 and were as follows:
counted by observers using three boats from 1975–2003
or an aircraft from 2004–2021. Maximum counts obtained Great White Pelican:
for Great White Pelican, White-breasted Cormorant, Grey- Y = 0.4424X + 155.4 (n = 15, r = 0.945, p < 0.001);
headed Gull and Caspian Tern at Lake St Lucia in each White-breasted Cormorant:
year surveyed are shown in Table 1. Y = 0.1146X + 66.291 (n = 10, r = 0.640, p < 0.05);
Records of breeding by these four species at Lake St Grey-headed Gull:
Lucia were obtained from the literature and unpublished Y = 0.2874X + 68.898 (n = 10, r = 0.854, p < 0.001);
information of EKZN Wildlife and are shown by year in Caspian Tern:
Table 2. Where available, estimates of the numbers of Y = 0.3343X + 55.833 (n = 13, r = 0.761, p < 0.01).
pairs of these species that bred are given but they seldom
reflect complete counts. The mid-points of reported ranges Future monitoring of numbers of individuals and of breed-
were used. Numbers of breeding adults were divided by ers will enable the usefulness of these relationships as
two to estimate pairs except for those at nests, which each predictors of breeding pairs to be assessed. Numbers breed-
represent a pair. Unattended chicks were assumed to ing at Lake St Lucia have been difficult to estimate because
represent pairs of Great White Pelicans, which only rear many colonies are sensitive to disturbance or difficult to
one chick, or halved and rounded upwards to estimate access. It is of interest that the best-fitting relationship was
pairs of White-breasted Cormorants, Grey-headed Gulls obtained for the Great White Pelican, which often breeds
and Caspian Terns, which all may rear two or more in large colonies and is the most accurately counted of
chicks (Hockey et al. 2005). The adjective ‘several’ the four species considered. The emergent use of drones
was conservatively taken to represent ‘two’, e.g. several may enable improved precision of breeding estimates for all
hundred Grey-headed Gull chicks in 1991 (Bowker and four species.
Downs 2012) to be 200 chicks and hence reflect 100 pairs. Counts of Great White Pelicans at Lake St Lucia were
Counts of African Penguins moulting feathers were made in 42 of the years between 1976 and 2021 (Table 1).
significantly linearly related to numbers that bred (Crawford
and Boonstra 1994). Therefore, for each of the four spe-
cies considered at Lake St Lucia the best-fitting linear
relationships between maximum counts of birds and
numbers of pairs breeding were derived. In years when
information was available for both these parameters, the
ratio between them was obtained. If the ratio was < 1 or
> 5 (> 12 for White-breasted Cormorant) the data were
excluded from the analyses. In the first instance (ratio <
1) it is not possible for the number of birds to be less than
the number of pairs, so the number of birds was likely
underestimated. In the second instance (ratio > 5) it is
probable that either breeding numbers were underestimated Great White Pelicans breeding at Bird Island, Lake St Lucia, in
or substantial proportions of birds did not breed at Lake 2015 (photo C Fox)

258