Extrapair Paternity and Breeding Synchrony in Gadwalls (Anas Strepera) in North Dakota

Abstract

Extrapair paternity and its correlates with breeding synchrony were examined in Gadwall (Anas strepera) using microsatellite DNA fingerprinting. Eleven of 261 ducklings (4.2%) within 8 of 29 broods (27.6%) had genotypes consistent with extrapair fertilizations, a comparable frequency to other species of waterfowl for which extrapair paternity has been documented. We found no evidence of intraspecific brood parasitism. The frequency of extrapair paternity was not significantly correlated with breeding synchrony. We suggest that female absences during egg-laying may have provided males with opportunities to pursue extrapair copulation when breeding was synchronous.

Resumen

Se examinó la paternidad extrapareja y su correlación con la sincronía reproductiva en Anas strepera utilizando huellas dactilares genéticas de ADN microsatelital. Once de 261 pichones (4.2%) en 8 de 29 nidadas (27.6%) tenían genotipos consistentes con fertilizaciones extrapareja, una frecuencia comparable a la de otras especies de aves acuáticas en las que se ha documentado paternidad extrapareja. No encontramos evidencia de parasitismo de cría intraespecífico. La frecuencia de paternidad extrapareja no se correlacionó significativamente con la sincronía reproductiva. Sugerimos que las ausencias de las hembras durante el período de postura podrían haber dado oportunidades a los machos para buscar cópulas con otras hembras cuando la reproducción era sincrónica.

SOCIALLY MONOGAMOUS BIRDS often copulate with partners other than their social mates (extrapair copulation, EPC), which can result in extrapair paternity (EPP) within broods. Breeding synchrony is an important demographic factor that appears to influence EPP rates. Across species of songbirds, breeding synchrony and EPP are positively correlated (Stutchbury and Morton 1995; Stutchbury 1998a, 1998b), but that association within species is not well understood. Traditional reasoning predicts that during synchronous breeding, most males in a population will be guarding their mates and, hence, unavailable for EPCs (Birkhead and Biggins 1987, Westneat et al. 1990). However, when few females are fertile simultaneously, more males are expected to be available for EPCs. A male-biased operational sex ratio (sensu Emlen and Oring 1977) should result in more males attempting EPCs with fewer females and a negative correlation between breeding synchrony and EPP (Birkhead and Biggins 1987, Westneat et al. 1990). That negative correlation has been documented in several species of birds (Reyer et al. 1997, Dunn et al. 1999, Saino et al. 1999, Thusius et al. 2001, Vaclav and Hoi 2002). Additional studies, however, have observed the opposite correlation (Stutchbury et al. 1997, Stutchbury 1998; Chuang et al. 1999; Zilberman et al. 1999), and that pattern may be expected when females seek and solicit EPCs (Stutchbury and Morton 1995). The association between breeding synchrony and EPP within a species may be influenced by the sex that seeks EPCs (see Thusius et al. 2001, Vaclav and Hoi 2002), but that idea has rarely been evaluated for a species in which females resist EPCs (but see Dunn et al. 1999).

Despite documentation of EPC in 55 species in 17 genera, there is no evidence that females of any species of waterfowl (order Anseriformes) solicit EPCs (McKinney and Evarts 1998). The Gadwall (Anas strepera) is a socially monogamous species of duck in which males make energetic pursuits of EPCs (Peters 2002). Males invest in their social pairbond by increasing their levels of vigilance during their mate's fertile period (Dwyer 1975) and behaving aggressively toward conspecifics (Titman and Seymour 1981, Peters 2002). On the basis of that observation, we expect that males are constrained by mate guarding (see Peters 2002), and that EPP will be negatively correlated with breeding synchrony in Gadwall. That association has been observed in Lesser Snow Geese (Anser caerulescens caerulescens), another species of waterfowl in which EPC activity is constrained by mate guarding (Dunn et al. 1999). However, some male ducks more actively seek EPCs during their mates' fertile period (Afton 1985, Sorenson 1994). Relative rates of EPCs in relation to their mates' fertile periods are unknown for male Gadwall, but males do seem to pursue EPCs when their mates are fertile. For example, while their mates were at the nest laying, males seemed to be active in pursuing extrapair females (Duebbert 1966, Peters 2002, also see Sorenson 1994). If males are more active in extrapair activities when their own mates are fertile, a positive association between breeding synchrony and EPP may be expected.

Extrapair paternity rates in waterfowl are of general interest because waterfowl differ from other bird taxa in two important ways. First, male waterfowl appear to use force to subdue and copulate with extrapair females (reviewed by McKinney and Evarts 1998). Second, male waterfowl have an intromittent organ that makes forced copulation feasible, and the morphology of that organ is related to sperm competition (i.e. mating strategy–EPC frequency; Coker et al. 2002). To date there are few estimates of EPP in waterfowl (Evarts and Williams 1987, Triggs et al. 1991, Larsson et al. 1995, Dunn et al. 1999, Eadie et al. 2000); this study adds to that sample. We also evaluate the presence or absence of intraspecific brood parasitism.

Methods

We sampled Gadwall families during the 1997 breeding season (15 June to 15 August) within the prairie potholes near Egeland, Towner County, North Dakota. Gadwall nests were located from late-April to mid-July by dragging a chain between two all-terrain vehicles through prairie nesting habitat and observing females as they flushed from the nest (M. Hoff and E. R. Loos unpubl. data). Thirty-seven females were subsequently captured using a long-handled hand net or with a nest-trap. A total of 284 eggs were collected from 29 nests and artificially incubated until hatching. We also sampled 12 adult males captured in decoy traps set in ponds. The femoral vein of each individual was punctured using a 3 cc syringe, and ~100 µL of blood was drawn with a capillary tube from each adult (n = 49) and 268 ducklings that hatched (6 to 11 ducklings per nest; mean = 9 ± 1.4 SD). Blood samples were stored in 1 mL anticoagulant buffer at –4°C. The DNA was extracted from blood samples using the InstaGene ^TM Genomic DNA Kit (Bio Rad, Hercules, California).

Five polymorphic autosomal microsatellite loci and two Z-specific loci (Table 1; Fields and Scribner 1997, Buchholz et al. 1998) were chosen for analysis. The forward primer of each pair was labeled with a fluorescent dye. Polymerase chain reaction was carried out using a 10 µL reaction mix following Fields and Scribner (1997) and Buchholz et al. (1998). Polymerase chain reaction conditions consisted of a 2 min denaturation at 94°C followed by 25 cycles (35 cycles for Sfiµ5 and Bcaµ10) of 30 s denaturation at 94°C, 30 s annealing at primer-specific annealing temperatures (Table 1), and 30 s elongation at 72°C, followed by a 5 min elongation at 72°C. Polymerase chain reaction products were analyzed using an ABI Prism ^®377 automated DNA Sequencer, and individuals were genotyped using GENESCAN ANALYSIS software (Applied Biosystems, Foster City, California).

Table 1.

Characteristics of primer pairs and polymerase chain reaction products of six microsatellite loci in 77 ^a adult Gadwall.

A total of 261 ducklings (97.4% of ducklings sampled; DNA extractions for seven ducklings failed) and their putative mothers (n = 29) were genotyped. To assess brood parasitism, genotypes of ducklings were compared directly with their putative mother. Extrapair paternity was evaluated by comparing genotypes of each duckling with their putative mother and brood mates. Paternal genotypes were inferred on the basis of the genotypes of the majority of the brood. Ducklings with novel alleles at a minimum of two loci (to allow for the possibility of mutations) and genotypes consistent with the putative mother were classified as extrapair young. The probability of excluding a putative parent as the real parent was calculated using a parentage exclusion probability (PE, Jamieson 1994; Table 1).

Breeding synchrony was evaluated on the basis of the laying dates of 138 Gadwall nests located in 1997 (see above; Fig. 1; M. Hoff and E. R. Loos unpubl. data). Eggs were candled to determine incubation stage (Weller 1956, modified by Delta Waterfowl Foundation for Gadwall), and laying dates were calculated by backdating. Breeding synchrony was calculated for the population using a synchrony index (SI; Kempenaers 1993), a measure of the average proportion of females that are fertile simultaneously throughout the breeding season. A modified version of that (proportion of females that were fertile during the fertile period of a given female, Yezerinac and Weatherhead 1997) was applied to individual females. Here, we define the fertile period of females to be the egg-laying stage to avoid making assumptions about the influences of sperm viability (but see Elder and Weller 1954 for Mallards [Anas platyrhyncos]) and sperm competition (last-male sperm precedence in Mallards, Cheng et al. 1983).

Fig. 1.

Number of females known to be laying (based on 138 nests; solid line, primary Y axis) and number of nests collected for paternity analysis (secondary Y axis) with (shaded bars; n = 8) and without extrapair young (open bars; n = 21). Each nest is plotted on the basis of the mid-date of egg-laying.

We compared the proportion of extrapair young within a brood to breeding SI and date of nest initiation using Spearman's . and the program JMP (SAS Institute 1997). Linkage disequilibrium and Hardy-Weinberg equilibrium were evaluated using ARLEQUIN 2.000 (Schneider et al. 2000).

Results

Eleven of 261 ducklings (4.2%) in 8 of 29 broods (27.6%) had novel alleles for at least two loci. Those 11 ducklings had genotypes consistent with the putative mother but inconsistent with the majority of the brood. Within broods, a low number of ducklings had genotypes inconsistent with full-sibling relationships (range = 1 to 3 extrapair ducklings per nest). Three additional ducklings had one novel allele each; and classifying those ducklings as extrapair young, 14 ducklings (5.3%) in nine broods (31.0%) were attributable to EPCs. On the basis of allele frequencies of adults and inferred paternal genotypes at six loci, the probability of excluding a putative sire or dam as a possible parent was 0.984 (Table 1). That probability was reduced to an average of 0.966 (depending on which locus was excluded), because ducklings were required to have novel alleles at two loci to be classified as extrapair.

The population-level SI was 26.3. There was no indication that SI and EPP were negatively correlated (Figs. 1 and 2). In fact, EPP exhibited a nonsignificant tendency (Fig. 2) to be more common in synchronous (SI > 25) than asynchronous (SI < 15) nests (6 of 17 [35.3%] vs. 2 of 11 [18.2%]; Fisher's exact test, P = 0.30). The lack of an association between SI and EPP persisted when we included the three ducklings with one novel allele each as extrapair young (r = 0.33, P = 0.08, n = 29). Also, the proportion of extrapair young per nest was unrelated to date of nest initiation (r = –0.02, P = 0.90, n = 29; see Fig. 1).

Fig. 2.

A plot of the proportion of extrapair young within each brood compared to breeding synchrony. There was no correlation between extrapair paternity and breeding synchrony (Spearman's . , r = 0.28, P = 0.14).

Discussion

Our results showed that only 4–5% of Gadwall ducklings in 28–31% of broods were extrapair young, and that EPP was not correlated with breeding synchrony. A negative relationship is expected when females do not seek EPCs and EPC activity by males is constrained by mate guarding (Birkhead and Biggins 1987, Westneat et al. 1990), and those prerequisites appear to be met in Gadwall (Peters 2002). Extrapair copulation activity by Lesser Snow Geese likewise is constrained by mate guarding, and Dunn et al. (1999) observed a negative correlation between breeding synchrony and EPP. The inconsistent associations between EPP and breeding synchrony in Gadwall and Lesser Snow Geese may result from different patterns of partitioning EPC and mate guarding behavior. Whereas male Lesser Snow Geese temporally partitioned mate guarding and EPCs by pursuing EPCs when their mates were nonfertile (i.e. incubating; Mineau and Cooke 1979), males of some species of ducks (e.g. Lesser Scaup [Aythya affinis] and White-cheeked Pintails [Anas bahamensis]) were more active in EPCs when their own mates were fertile (Afton 1985, Sorenson 1994). Male White-cheeked Pintails diurnally partitioned EPCs and mate guarding by pursuing EPCs primarily in the mornings (ducks tend to lay eggs in the morning) when their mates were at the nest laying (Sorenson 1994), and Gadwall appear to fit that model of diurnal partitioning (Duebbert 1966, Peters 2002). As a result, male Gadwall were available for EPCs when their own mates were fertile and breeding was synchronous, and that availability may have resulted in the lack of an association between breeding synchrony and EPP.

Male Gadwall likely had opportunities to attempt EPCs with fertile females while their own mates were in the process of egg-laying because individual females' laying schedules vary over the course of producing a clutch. Not all females lay at the same hour of day, and that provides opportunities for males with fertile mates to encounter additional fertile females while their own mates are tending to the nest. For example, Duebbert (1966) observed that female Gadwall arrived at the nest to lay each morning between the hours of 0500 and 0700 CST. Furthermore, as laying progresses, female Gadwall spend an increased amount of time on the nest (E. R. Loos pers. comm.). That variability may result in the availability of some fertile females while some males are temporarily emancipated from mate guarding, and males could have used that opportunity to pursue EPCs when breeding was synchronous. In fact, synchronous nests were twice as likely to contain EPP than were asynchronous nests (35 and 18%, respectively). Although that trend was not significant, it is consistent with the idea that male Gadwall were more active in EPCs when their own mates were fertile. Behavioral observations of EPC activity (in relation to their mates' fertile status) by individual males are needed to address that possibility.

The frequency of EPP in Gadwall is comparable to frequencies in other species of waterfowl for which extrapair paternity has been reported. For example, 3% of Mallard ducklings, 2 to 5% of Lesser Snow Goose goslings, and 2% of Ross's Goose (Anser rossii) goslings were extrapair young (Evarts and Williams 1987, Lank et al. 1989, Dunn et al. 1999). However, EPP appears to be absent in Blue Ducks (Hymenolaimus malacorhyncos;Triggs et al. 1991), Barnacle Geese (Branta leucopsis; Larsson et al. 1995), and Barrow's Goldeneye (Bucephala islandica; Eadie et al. 2000). In comparison to those species of waterfowl, EPP is more common in passerines, in which 18 ± 17% of offspring were extrapair (Westneat and Sherman 1997, Wink and Dyrcz 1999).

Intraspecific brood parasitism has been observed in 4–34% of Gadwall nests under high nesting densities (reviewed by LeSchack et al. 1997). However, in this study, all ducklings had genotypes consistent with their putative mothers. Given a 97% probability of detection, there was no evidence of brood parasitism in that population. Brood parasitism in Gadwall is generally attributed to two females choosing the same nest site under high nesting densities (on islands or dikes) rather than true parasitism (LeSchack et al. 1997). Nests sampled here were dispersed throughout upland nesting habitat, and that distribution may have limited the occurrence of brood parasitism. However, parasitism between close relatives and low frequencies of parasitism may not have been detected.

Acknowledgments

Many people have helped immensely during the course of this study. F. Rohwer offered valuable comments during the development of this study; K. Scribner provided microsatellite primers for initial screening; C. Seal, D. Evelsizer, and J. Mock provided field help; and B. Razvi, P. Ford, S. Julian, D. Rarus, and J. Julian helped with laboratory work. We thank the Department of Biology, Frostburg State University, the College of Veterinary Medicine, Virginia Polytechnic Institute and State University, and the Aquatic Ecology Lab, U.S. Geological Survey, for the use of facilities to complete genetic analyses; and we especially thank L. Wiegt and T. King for their guidance and support. M. Hoff and E. R. Loos provided the critically important data on Gadwall nesting chronology. J. L. Hoogland, K. E. Omland, C. R. Feldman, B. Kondo, T. E. Murphy, and two anonymous reviewers offered valuable comments on this manuscript. Delta Waterfowl Foundation provided financial support, and Frostburg State University provided Teaching Assistantships and lab supplies during the course of this study. K. Omland provided additional support, space, and guidance during the completion of this study. We would like to dedicate this manuscript in memory of Frank McKinney.