Leslie G. Underhill

Growth and its relationship to fledging success of African black oystercatcher Haematopus moquini chicks

We investigated the growth of African black oystercatcher Haematopus moquini chicks on Robben Island, South Africa, over three austral summers, 2001-2004. Using a robust regression analysis to determine the growth parameters of chicks of known and unknown age we found that oystercatchers from our study population had a Gompertz growth rate coefficient that was 2% less than predicted for body mass based on the equation for waders. Leg growth lagged initially, then increased and slowed again as the chicks became older, whereas wing growth was slow initially but increased with age. Chicks with small growth rate coefficients for body mass exhibited retarded growth of all body measures except wing length. This enabled these chicks to fledge in a shorter period of time than their slow growth would otherwise allow. The growth rate of body mass was observed to vary greatly between chicks. Fast-growing African black oystercatchers had a shorter pre-fledging period; were larger at fledging and were more likely to fledge successfully. African black oystercatchers display sibling rivalry, and once a dominance relationship is established, the larger chick remains so during the pre-fledging period. Larger siblings fledged earlier and at a heavier mass than the smaller siblings and this may improve their chances of survival. Neither hatching date nor brood size influenced the growth rate coefficients.

Results of the 2001 census of African penguins Spheniscus demersus in South Africa: first measures of the impact of the Treasure oil spill on the breeding population

The Treasure, a bulk ore carrier, sank between Dassen and Robben islands on 23 June 2000. Subsequently, more than 19 000 African penguins Spheniscus demersus were oiled. A further 19 500 unoiled penguins were caught at Dassen and Robben islands and relocated to Cape Recife in the Eastern Cape to prevent them from becoming oiled. Despite concerns about the negative impacts of the incident on the penguin breeding population, especially on Dassen and Robben islands, the number of penguin breeding pairs in South Africa increased by 14% between 2000 and 2001. The increase appears to be related to an abundance of the two most important prey items of the African penguin, anchovy Engraulis capensis and sardine Sardinops sagax, in 2001. A larger proportion of clean penguins that were relocated to Port Elizabeth bred in 2001 than penguins that were oiled and cleaned. The results highlight the importance of oiling and rehabilitation and availability of food, especially of anchovy and sardine, for the future conservation of the African penguin.

BREEDING, MASS AND PRIMARY MOULT OF EUROPEAN STARLINGS STURNUS VULGARIS AT DASSEN ISLAND, SOUTH AFRICA

Summary Cooper, J. & Underhill, L.G 1991. Breeding, mass and primary moult of European Starlings Sturnus vulgaris at Dassen Island, South Africa. Ostrich 62: 1–7. European Starlings Srurnus vulgaris breed from September to November at Dassen Island, South Africa. Mean clutch size was 3,8 and mean brood size 2,8. An estimated 0,7 juvenileslbreeding attempt survived to 2–3 months of age. Winter (post moult) masses were higher than in summer in juveniles, but not in adults. Adult and juvenile starlings undertook a comlete post-breeding moult. Based on percentage feather mass grown, primary moult took 108 d in aduis and 106 d in juveniles. On average, juveniles commenced moult 19 d later than did adults.

The coefficient of variation biplot

This note introduces the coefficient of variation biplot, and suggests that it will provide a useful graphical display of data matrices in which the relative variability of the columns is of interest.

AN ANALYSIS OF THE MORTALITY AND SURVIVAL OF HARTLAUB'S GULL

Summary Underhill, L. G. & Underhill, G. D. 1986. An analysis of the mortality and survival of Hartlaubs Gull. Ostrich 57: 216–223. Ringing recoveries of Hartlaubs Gulls Larus hartlaubii ringed at Robben Island and Rondevlei are analysed. The variation in the recovery rate between cohorts of chicks ringed on Robben Island on the same day and between years is attributed to initial mortality (mortality after ringing but before reaching the mainland). Estimates of initial mortality are negatively correlated with total rainfall (a measure of storminess) during the breeding season, indicating that stormy weather contributes to breeding success, presumably through increasing food availability. The survival probabilities of first-year, second-year and adult Hartlaubs Gulls are estimated, and found to be independent of year.

Leach's storm petrels Oceanodroma leucorhoa off southern Africa : breeding and migratory status, and measurements and mass of the breeding population : research paper

Leachs storm petrel Oceanodroma leucorhoa have bred at Dyer Island since at least 1995; a review of earlier records in the light of this discovery indicates that breeding has almost occurred since at least the 1970s. The South African offshore islands on which Leachs storm petrel almost certainly breeds are Jutten, Dassen, Dyer and St Croix. The South African population is probably of the order of 25 pairs. The migrant component of the Leachs storm petrel population visiting southern Africa is mainly confined to deep ocean waters (generally 2000–5000 m deep) and occurs in particularly high densities around Agulhas Rings where ring-edges apparently provided excellent foraging opportunities. Leachs storm petrels are comparatively scarce over the shelf areas (defined as coastal water of less than 2000 m depth) of Namibia and western South Africa. Given their abundance at sea, their distance offshore, the seasonality of their occurrence and their moult status, these birds cannot be part of the South African breeding population, and must therefore be mainly from colonies of the north Atlantic Ocean. The Leachs storm petrels breeding in South Africa appear to be relatively large, comparable in size to populations breeding in the north Atlantic Ocean. Molecular techniques should be used to determine the most closely related population, and possibly the period for which the South African population has been isolated.

Knots (Calidris canutus) from Germany and South Africa parasitised by trematode Cyclocoelum mutabile.

Two Knots, among 88 dissected, contained the trematodeCyclocoelum mutabile. One represents the first record of this trematode from the Afrotropics. Both birds showed retardation of moult, and it is possible that such parasites have an indirect lethal effect, and that they affect breeding performance of this migrant wader with a tight annual schedule. Zwei von 88 sezierten Knutts enthielten den SaugwurmCyclocoelum mutabile. Einer davon stellt den Erstnachweis dieses Trematoden für die Afrotropis dar. Beide Vögel waren in ihrer Mauser verspätet. Es ist möglich, daß die Parasiten indirekt tödliche Auswirkungen haben und den Brutverlauf des Knutts beeinflussen, dessen Jahreszyklus sehr gedrängt ist.

Impact of Climate and Predation on Autumn Migration of the Curlew Sandpiper

Abstract. Using constant-effort catch data, causes of annual variation in the timing of migration of Curlew Sandpipers (Calidris ferruginea) migrating through Ottenby, Sweden, as well as the trend in timing of migration from 1946–2005, was investigated. Variation in the timing of autumn migration of adult and juvenile Curlew Sandpipers was influenced by breeding success connected to predation pressure on the Arctic breeding grounds. Median migration date of adult birds was significantly later in good breeding years compared with poor breeding years while the migration of juveniles was earlier in good breeding years compared with poor breeding years. Also, adults migrated earlier when the average temperature in June was warmer. Median migration dates of adults have advanced by 23 days from 1946–2005, but the migration dates of juveniles have remained unchanged. Unchanged migration dates of juveniles indicate that earlier departure of the adult Curlew Sandpipers from the breeding grounds was not due to earlier breeding. Evidence suggests that declining breeding productivity as a result of increasing predation on broods of shorebirds might, over the years, be the reason for the observed pattern of early departure of adults from the breeding grounds. One possible consequence of earlier migration is a mismatch between timing of migration and periods of food abundance on migration routes and at the wintering grounds, leading to a decline in adult and juvenile survival and population size.

Primary moult, mass and movements of the Rock Pigeon Columba guinea in the Western Cape, South Africa

Underhill, L.G. & Underhill, G.D. 1997. Primary moult, mass and movements of the Rock Pigeon Columba guinea in the Western Cape, South Africa. Ostrich 68 (24): 86–89. Rock Pigeons Columba guinea in the Western Cape, South Africa, take an estimated 7.2 months to complete primary moult. The mean starting and completion dates are 26 December and 2 August, with 95% of birds starting and completing within two months of these dates. The overall mean mass was 344 g, but birds were heaviest in winter (356 g) and lightest in spring and summer (334 g). Twenty-four of 48 recoveries of Rock Pigeons ringed in the Western Cape were more than 2 km from the ringing site. These recoveries demonstrate movements of Rock Pigeons between the mountains of the Cape Peninsula and the wheat-growing areas to the northeast.

Breeding phenology of African Black Oystercatchers Haematopus moquini on Robben Island, South Africa

The timing of the start and duration of breeding and the effect of these on breeding productivity were analysed for African Black Oystercatchers Haematopus moquini on Robben Island, South Africa, over three breeding seasons from 2001 to 2004. African Black Oystercatchers have a long breeding season, from November to March, enabling a second or sometimes a third nesting attempt after depredation or a natural disaster that destroys nests. The incubation starting date was known or estimated for 205 African Black Oystercatcher clutches over three breeding seasons. Using the estimated starting date of incubation for each nest we found that the 2001/02 and 2002/03 breeding seasons started approximately two weeks earlier than the 2003/04 season and that the 2002/03 breeding season ended approximately one month later than the other two. The late start of the 2003/04 breeding season may be a consequence of greater predation compared to the previous two breeding seasons. The 2003/04 breeding season was the shortest and had the lowest breeding productivity. The 2002/03 breeding season had a lower breeding productivity than 2001/02 because of the loss of a large number of nests during storm surge and spring high tides. The start and duration of the egg-laying period are influenced by predation risk and weather conditions.