Timing of breeding and reproductive costs in Collared Flycatchers

DAVID A. WIGGINS,13 TOMAS PART,2 AND LARS GUSTAFSSON'

Many recent studies of birds have attempted to measure a cost of reproduction by correlating current investment with subsequent parental survival and reproductive performance (see Dijkstra et al. 1990, Roff 1992, Stearns 1992). Demonstrating a link between current and future reproduction provides an important test of within-generation life-history tradeoffs (Williams 1966, Gadgil and Bossert 1970, Charnov and Krebs 1974, Daan et al. 1990, Stearns 1992). Studies of birds have typically relied on experimental manipulations of brood size (e.g. Gustafsson and Sutherland 1988, Nur 1988) because brood size often is correlated with the amount of care provided by parents and with the number of offspring subsequently recruited into the population.

Timing of breeding is another factor that may be significantly correlated with parental survival (e.g. lower survival for late breeders; Harvey et al. 1988, Wiggins 1991, Verhulst et al. 1995) and offspring recruitment (e.g. decreased recruitment with later breeding; Price et al. 1988). Although recent experimental work has demonstrated a direct link between the timing of breeding per se and annual reproductive success (Verhulst and Tinbergen 1991, Brinkhof et al. 1993, Norris 1993), such studies have not assessed the long-term effects of breeding-date manipulations on subsequent parental survival or reproduction. If breeding date per se is negatively correlated with parental survival or future reproduction (e.g. owing to time constraints related to initiation of molt or migration), then experimental delays in breeding should result in reduced parental survival or reproduction in subsequent years. The limited available evidence is equivocal: Verhulst et al. (1995) found that female quality, but not breeding date, was responsible for the seasonal trend of decreasing female survival in Great Tits (Parus major), whereas among Blue Tits (P. caeruleus), delayed breeding led to poor adult survival and lower future reproductive success (Nilsson and Svensson 1996).

Experimental manipulations of breeding date are difficult to achieve without inducing concurrent effects on parental body condition. The two most common methods used to alter breeding date, clutch removal and temporary clutch replacement with dummy eggs, may result in considerable increases in energetic effort by males (in species where incubation feeding is common, such as Ficedula flycatchers) and females (via increased incubation costs; e.g. Moreno and Carlson 1989). Although previous studies have typically concluded that such experimental effects are minimal (e.g. Verhulst and Tinbergen 1991, Brinkhof et al. 1993, Nilsson and Svensson 1996), one of the primary concerns in our study was to delay the timing of hatching while minimizing the energetic consequences (potentially induced by prolonging the incubation period) for males and females.

In this study, we attempt to link the timing of breeding per se with the subsequent survival and reproductive success of adult Collared Flycatchers (Ficedula albicollis). Thus, in contrast to most previous studies of reproductive costs in birds, our approach was to manipulate the timing (by extending incubation), rather than the level (by changing brood size) of parental care. However, because offspring production is correlated with timing of breeding, we also consider how the interaction of timing and brood size may affect the future survival and reproduction of parents. If later breeding per se entails significant costs to the parents (e.g. via decreased time available for molt prior to migration, and / or increased overlap between molt and breeding), then a delay in breeding date may have significant effects on subsequent parental survival and/or reproduction (cf. Nilsson and Svensson 1996).

Study area and methods.-The study was carried out in 1992 and 1993 on the island of Gotland (5710'N, 1820'E) in the southern Baltic Sea (see Part and Gustafsson 1989 for description of study site). Flycatchers return to the island in early May, lay eggs from mid-May to mid-June, and leave the island in July and August. To determine clutch initiation dates, we checked nest boxes every two to three days from early May to early June. During the nestling phase, we measured and banded most adults and all nestlings.

In 1992, we delayed hatching by seven days in 108 randomly chosen experimental nests. Clutches were removed on the (presumed) morning of clutch completion, placed in a refrigerator at 7C, and replaced by an equal number of artificial eggs. Nests were rechecked for a new egg on the day (or days) following clutch removal. If a new egg was present, it was added to the original clutch and replaced with a dummy egg. Clutches were replaced in their original nests seven days after clutch completion, thus extending the incubation period from approximately 13 to 20 days. In 44 of the experimental clutches, one egg was removed to mimic the seasonal decrease in clutch size (see Wiggins et al. 1994a).