Behavioral Ecology Behav Ecol Sociobiol (1985) 17:363-369 and Sociobiology Springer-Verlag 1985 Parental energy expenditure: a proximate cause of helper recruitment in the pied kingfisher (Ceryle rudis) Heinz-Ulrich Reyer 1, and Klaas Westerterp 2 1 Max-Planck-Institut fiir Verhaltensphysiologie, D-8131 Seewiesen, Federal Republic of Germany 2 Biomedisch Centrum, Postbus 616, NL-6200 MD Maastricht, The Netherlands Received October 15, 1984 / Accepted February 25, 1985 Summary. Energy expenditure of adult Pied rect decision they require proximate mechanisms Kingfishers was measured with doubly-labeled which are (1) constantly available and (2) good water. Results were related to reproductive success predictors of fitness, i.e., they must be closely re- of parents aided and unaided by helpers. Energeti- lated to current reproductive success and the prob- cally stressed parents in a colony with poor food ability of survival. Energy balance is a likely candi- supply accepted potential helpers more often than date. Although a negative relationship between en- unstressed birds in another colony where food was ergetic stress in one year and survival into the next easily available. This treatment of helpers was re- has only been shown for a few species (Nur 1984), versed in both colonies through experimental ma- including the Pied Kingfisher (Reyer 1984), current nipulation of clutch size and hence energetic stress. reproductive success in many species is closely Results are discussed in relation to the costs and linked to energy balance and body condition benefits that helpers incur on the parents' fitness. (Drent and Daan 1980). These in turn depend on environmental conditions, particularly food. The conditions in a given season determine whether a bird will breed at all, how early and how often eggs will be laid, how big the clutch will be, how Introduction much food the parents can bring and consequently In recent years several authors have attempted to what the growth and survival rate of the young explain cooperative breeding in birds as an adapta- will be. tion to ecological conditions (e.g. Brown 1974; Or- It is conceivable that the parents' decision to ians et al. 1977; Gaston 1978 ; Koenig and Pitelka accept helpers or not also depends on energy bal- 1981 ; Emlen 1982, 1984). The critical test for such ance and body condition. Perhaps a bird could a hypothesis is to compare inclusive fitness values compare the food requirements of the young (as of conspecifics pursuing different behavioral stra- communicated in their begging) to its own feeding tegies in identical environments and of those pur- capacity and the amount of competition from help- suing identical strategies in different environments. ers. This hypothesis combines conclusions from the Such a comparison is possible in the Pied above mentioned energetic studies with an idea, Kingfisher (Ceryle rudis). It was shown previously originally suggested by Orians et al. (1977) and re- that males choose the strategy which yields the cently modelled by Brown (1982, 1984) and Emlen highest inclusive fitness under the prevailing eco- (1982), namely that parents should accept or reject logical and demographic conditions (Reyer 1980, helpers according to need. We tested this hypothe- 1984). For male breeders this choice means to re- sis by measuring first the daily energy expenditure ject potential helpers (which are always males and (DEE) of feeding adult Pied Kingfishers in two therefore also potential rivals) when food condi- ecologically different colonies, using doubly-la- tions are good but to accept them when conditions beled water. We then related the DEE of parents are too poor for parents to raise all their hatchlings to their reproductive success, their behavior to- unaided. If birds are to make this ultimately cor- wards potential helpers and the begging duration of their young under normal conditions. Finally * To whom offprint requests should be sent we manipulated clutch size, changing the begging 364 of young and the energetic stress of parents. We 600- only consider the treatment of "secondary help- o o/ ers", unrelated to the breeders, and not that of ~ / related "primary helpers". The latter apparently 500- , / do not compete with the breeders for females and were tolerated under all observed circumstances X (Reyer 1980, 1984 and in preparation). _~ 4o(> o / & Methods E / "~ 300- / The study was carried out between 1981 and 1983 on two breed- .9o ing populations at Lake Victoria and Lake Naivasha, Kenya .................... 2~ o (Reyer 1980). Adult birds who fed nestlings were caught in / o :i I the late afternoon, weighed and then injected with 0.5-0.8 ml .~ 200- /g o" of 18% (excess atom percent) oxygen-18 enriched and 10% deuterium enriched water (ZH2tSO). After 1 h of equilibration o o / ": "j ca. 60 ~tl blood was extracted from the wing vein before releas- ........... _/__ ..... ing the bird. Birds were recaptured and weighed, and blood 100" _. " samples taken again 24 h (or a muttiple thereof) later. CO 2 production was calculated from the reduction of 2H and a aO between the first and second sample as described previously 0-,/. ~ ..//f/"o" ~ ~ (Westerterp and Bryant 1984) and then converted into DEE 130 160 1~0 210 2;0 2~0 (kJ) with an RQ value of 0.8 for fish protein on which the energy expenditure [ kJ/d ] birds feed almost exclusively. Additionally, the number, type, Fig. 1. Amount of food delivered to nestlings (kJ/ad*d) in rela- and size of fish provided for nestlings by the injected bird was tion to daily energy expenditure of feeding adults (kJ/d). monitored during the whole activity period. The daily food o---o=Lake Naivasha, y=4.693x--718.3; n=17, r= contribution of the 2HztaO-injected bird was calculated from 0.756, P<0.001. (e, o=Lake Victoria, y=l.485x-- these data as describe~t earlier (Reyer 1984) by using the length- 209.9; n = 15, r=0.545, P= 0.036. - ...... Upper limit of energy weight relationships in Table 1. In the same way the contribu- expenditure (from Fig. 2) and resulting feeding capacities re- tions of all other birds feeding at the particular nest were re- corded. From the total amount of food and the known number spectively of nestlings the average energy intake/nestling and day was calculated. The effect of that intake on growth was measured a lower energy yield per fish at Lake Victoria which by weighing nestlings before and after the experiment when must be compensated for by catching more fish in the post-absorptive state. Weights were usually taken at 530 h, before the first feed of the day. (cp. allometric equations in Table 1), (2) greater The nestlings', hunger was monitored by placing a micro- turbidity of the water (leading to lower hunting phone at the entrance of the nest chamber and connecting it success) and longer distances between fishing to a tape recorder in the observation hide. Ca. 10 s recordings grounds and the breeding colony at Lake Victoria were made every 5 min throughout the day to determine wheth- resulting in more flying and hovering time per er the chicks were begging. Begging duration was expressed as the percentage of recordings with begging, excluding cases catch. in which the nestlings were being fed or had been fed in the The ranges and averages of body mass before previous minute. All the above measurements on adults and injection were not significantly different between young were taken when the nestlings were 5 to 17 days old. birds from the two lakes (both medians 76 g; Lake In Pied Kingfishers this is the period of linear growth (Reyer, unpunished data). Details of the manipulation experiments will Victoria: interquartile range 72.0-80.0, n=15; be mentioned with the results. Lake Naivasha: interquartile range 70.5-78.5, n = 17; z--0.566, N.S., Mann-Whitney U-test, two- tailed). Results Birds with a DEE of up to 210 kJ maintain or even increase their body mass (Fig. 2). Above Energy expenditure and feeding capacity that they increasingly lose mass (P < 0.001, Mann- As the daily energy expenditure of adults increases, Whitney U-test, one-tailed), as much as 9.2%/day. the amount of food delivered to nestlings rises in Thus 210 kJ seems to represent some physiologi- a linear fashion at both lakes but with significantly cally determined threshold which cannot be ex- different slopes (Fig. 1 ; P = 0.037, analysis of co- ceeded for prolonged periods without a reduction variance; Sokal and Rohlf 1969). Thus, a Lake in body mass. Possible reasons for this reduction Victoria bird will achieve a lower feeding contribu- will be addressed in the Discussion. Entering the tion than one from Lake Naivasha for the same threshold of 210 kJ into Fig. 1, we can predict that amount of energy expended. The ecological rea- on average a pied kingfisher at Lake Naivasha can sons for this difference include (Reyer 1980): (1) deliver as much as 267.2 kJ/day without losing 365 Table l. Relationship between standard length sl (mm), dry mass M e (g) and energy content (kJ/g Md) for cichlid fishes and Engraulicypris argenteus (Cyprinidae) from Lake Victoria and for Tilapia ssp. (Cichlidae) and Micropterus salmonides (Centrarchi- dae) from Lake Naivasha. r = Pearson correlation coefficient between sl and M d (sample size = n x). x and SD = mean and standard deviation of energy content/g Md as analyzed by bomb calorimetry (sample size=n2) Prey type Regression r n~ x SD n 2 Engraulicypris log Me = 0.037" s/--2.320 0.98 52 19.95 0.61 3 Cichlidae log M d= 0.026* sl- 1.461 0.99 95 18.89 0.23 3 Tipapia ssp.