Effects of food supplementation on the timing of nest initiation in belted kingfishers

Ecology, Dec, 1997 by Jeffrey F. Kelly, Beatrice Van Horne

INTRODUCTION

The timing of initiation of nesting is related to reproductive output in several bird species (reviewed in Martin 1987:460). Numerous factors contribute to variation in the timing of reproduction among individuals, including age, fat reserves, habitat quality, availability of mates, and time of arrival on the breeding ground. Probably because of the pioneering work of Lack (1954:53-66) and Perrins (1970), most investigations of variation in timing of reproduction have focused on the effects of food availability. Lack proposed that birds use food availability as a proximate cue to time nesting so that they will be feeding nestlings and fledglings during periods of peak food abundance. Perrins suggested that the ability of females to acquire food resources for egg-laying could constrain the onset of reproduction directly. Although there are differences between the hypotheses of Lack and Perrins, they both assume that energy availability is the primary source of variation in the timing of nest initiation. Because these hypotheses make similar predictions and are based on energy limitation, we refer to them jointly as the energy constraint model.

The energy constraint model has been tested primarily through food supplementation experiments. The primary prediction of this model, that birds receiving supplemental food will begin reproduction earlier than those not receiving food, has been upheld by the majority of studies (reviewed in Arcese and Smith 1988, Boutin 1990). The agreement between the results of food supplementation experiments and this prediction of the energy constraint model has led ornithologists to conclude that food is the primary proximate factor affecting the timing of nest initiation (e.g., Arcese and Smith 1988, Daan et al. 1988, Nilsson and Svensson 1993).

No study has examined the effect of supplemental food on the timing of reproduction in piscivorous birds, although Davis and Graham (1991) found that the reproductive performance of the piscivorous Green Kingfisher (Chloroceryle amazona) was affected by food availability. Because most food supplementation studies have provided seeds, mealworms, or carrion (Boutin 1990), and patterns of abundance and availability of fish are likely to be broadly different from these types of food, food supplementation studies of fish-eating birds have the potential to broaden the inferences drawn from prior studies. For these reasons, we supplemented food of piscivorous Belted Kingfishers (Ceryle alcyon) to determine whether the timing of their reproduction was constrained by available energy. Our results led us to develop an alternative to the energy constraint model for the relationship between food availability and timing of reproduction in birds.

METHODS

We worked along the Cache la Poudre River in Fort Collins, Colorado, USA. In this region of northern Colorado, the breeding biology of Belted Kingfishers was similar to that reported elsewhere (review by Hamas 1994). Belted Kingfishers nested in steep, unvegetated dirt banks. They dug burrows that were [approximately] 1 m in length and laid 5-7 eggs. Belted Kingfishers were socially monogamous and raised only a single brood per season. As is typical in the interior northern portion of this species' range, males in our study area were generally resident, whereas most females left the study area during winter (there were 5-10 males per female during winter). The density of kingfishers was about one nesting pair per 1.5 km of river; males were strongly territorial year-round. Migrant females began returning to the study area in late March and early April.

We located and monitored Belted Kingfisher nests along the Cache La Poudre River during the 1992 through 1995 breeding seasons. We searched 20 km of river for nests weekly from March through June. Nests were located by examining river banks for burrows, as well as by following adult kingfishers. During 1992, we found most nests during the incubation stage; during 1993-1995, we were more familiar with the study area and found most nests during the burrow construction phase. After locating a nest, we watched it every fourth day for 1-2 h and recorded all behaviors of adults, including food deliveries. We estimated dates of egglaying and hatching, based on observations during these nest watches. Early in the nesting cycle, female kingfishers spent a large portion of their time perched outside their burrow and were fed by their mate. By back-dating at successful nests, we found that the period when females were perched outside their nest was coincident with the egg-laying period. based on this pattern, we used the presence of the female perched outside the burrow for long periods as an indication of egg-laying. After incubation began, adults were rarely seen near the burrow except when entering and leaving. Using information from successful nests, we were also able to determine that the adults became noisy when the eggs hatched, and frequently brought food to the burrow. Thus, it was clear from changes in adult behavior when a change in the nesting cycle occurred. Because of concern over disturbance, however, we did not confirm these observations by excavating nests during egg-laying or incubation. We used the midpoint between two observation periods as an estimate of the date that the first egg was laid. We refer to this date as the nest initiation date. When necessary, we backdated to estimate nest initiation date by assuming that kingfishers laid one egg per day and incubated for 22 d after the final egg was laid (Hamas 1994).