2001. Do southern African songbirds live longer than their European counterparts? -Oikos 93: 235 -249.Understanding life history evolution in tropical and southern hemisphere birds has been hindered by a paucity of knowledge of key life history traits and this is particularly true of African songbirds. Here we use a unique long-term mark-recapture data set collected over 16 years in Malawi (latitude 16°S) to estimate adult survival rates for 28 African passerine species. Survival of these and 11 other African songbirds (taken from the literature) showed a bi-modal distribution with annual survival of insectivores and nectarivores (bulbuls, thrushes, warblers and sunbirds) averaging 72% (quartiles 63 -80%) compared to 54% (50 -62%) in granivores (weavers, finches and canaries). The mean adult life expectancy of African insectivores and nectarivores (3.1 yr) was more than twice that of related European insectivores (1.4 yr) and nearly twice that of African granivores (1.6 yr). These marked differences in survival were highly significant after controlling for body mass and phylogeny. Among African songbirds there was a strong negative correlation between adult survival and clutch size with granivores laying relatively large clutches and living relatively short lives. We hypothesize that these differing life history trade-offs reflect variation in the seasonality of food resources whereby survival rates of northern temperate songbirds may be limited by food availability and cold weather during winter, while survival of southern African granivores may be limited by the influence of a variable and unpredictable rainfall regime on seed availability.

No abstract

No abstract

There are many methods of calculating mortality rates in birds, but none appear to be entirely satisfactory, due to the number of possible instances of bias when dealing with living organisms. The method used here is that used by Brown & Pomeroy (In press, Proc. 5th Pan-Afr. Om. Congr. ), for their attempt to show that mortality rates are related to weight. It is a very simple method, but since the samples obtained from recaptures are very small, there is no justification for the use of a more sophisticated method.Using recapture data, the number of birds surviving each successive year after ringing is listed. Then the numbers surviving at hypothetical constant annual mortality rates are calculated, to see which rate best fits the actual figures. This method assumes that birds do not lose their rings, emigrate or otherwise escape capture and that the annual mortality rate of adults is constant. None of these are necessarily true. In their middle years, bird mortality rates do not appear to differ greatly, but without supporting evidence this remains an assumption and Fry (1980, Proc. 4th Pan-Afr. Om. Congr.: 333-343.) ~ ~~ ~ ~~ Annual Predicted survival with annual mortality mortality rate of rate (%) -----37% I 36% I 35% 1 34% I 33% TABU 1 NUMBER OF ADULT LITTLE BEE-EATERS RINGED AND RECAUGXr OVW YEARS, "HEIR ANNUAL MORTALITY RATES AND THE NUMBER W W G AT CERTAIN RYPOTHBTICAI, CONSTANT ANNUAL MORTALITY RATES. 5 6 I --suggests that some species show a decrease in mortality rate with age. One hundred and twenty-five apparently adult and 42 visibly immature Little Bee-eaters Merops pusillus were ringed between March 1974 and June 1979 at Nchalo, Mala+i (16 16s; 34 55E). Some of these were recaptured in succeeding years, up to January 1982. Years were taken from July to June.As immature Little Bee-eaters cannot be identified after they are four months old, there are probably many immatures included in the adult group. Table 1 gives the number of apparent adults ringed and recaught over seven years (none has been recaught more than seven years after being ringed). Table 2 gives the number of known immatures ringed and recaught over six years. The adult mortality rate in the first year is 76 %. This is almost certainly mainly due to immature mortality and dispersal, as the immature first-year mortality rate is similar (7479, although adults may be partly migratory, which could account for a small proportion of the loss. After the first year the percentage mortality is considerably less, so that in calculating the hypothetical annual mortality rates for comparison with the actual figures, the first year was excluded. Five hyopthetical rates were calculated (to two decimal places, but rounded off to the nearest whole bird) TABLE 2 N U M B E R OF IMMATURE LITTLE BEE-EATERS RINGED AND RECAUGHT OVER SIX YEARS, THEIR ANNUAL MORTAIXIT RATES AND THE NUMBERS SURVIVING AT CERTAIN HYPOTHETICAL CONSTANT ANNUAL MORTALITY RATES. Predicted survival with annual mortality rate of year I rate(%j -----Year I (actual)