Triparental care in the collared flycatcher (Ficedula albicollis): Cooperation of two females with a cuckolded male in rearing a brood

Abstract Certain predominant forms of mating and parental care systems are assumed in several model species among birds, but the opportunistic and apparently infrequent variations of “family structures” may often remain hidden due to methodological limitations with regard to genetic or behavioral observations. One of the intensively studied model species, the collared flycatcher (Ficedula albicollis), is usually characterized by social monogamy with polyterritorial, facultative social polygyny, and frequent extrapair mating and extrapair paternity. During a brood‐size manipulation experiment, we observed two females and a male delivering food at an enlarged brood. A combination of breeding phenology data (egg laying and hatching date), behavioral data (feeding rates) from video recordings at 10 days of nestling age, and microsatellite genotyping for maternity and paternity suggests a situation of an unrelated female helping a pair in chick rearing. Such observations highlight the relevance of using traditional techniques and genetic analyses together to assess the parental roles within a population, which becomes more important where individuals may dynamically switch from their main and presupposed roles according to the actual environmental conditions.

Breeding and parental roles in one of the most investigated model species in behavioral ecology, the collared flycatcher (Ficedula albicollis), have seemed to be only moderately plastic. We have been studying a population since the early 1980s (see below for details).
At our study site, these birds arrive from spring migration in late April. During a breeding season, females lay one clutch with 5-8 eggs incubated exclusively by them, but nestlings are fed by females and males together. In our study population, only 5.7% of males become socially polygynous during their lifetime (Herényi et al., 2014), based on capture data. Due to female aggression (Hegyi et al., 2007), polyterritoriality typically occurs in polygynous males with a territory distance of 15-300 m. In contrast to the very low rate of social polygyny, experimental mating restriction in combination with sperm counting on the perivitelline layer (Michl et al., 2002) and microsatellite-based paternity analyses (Rosivall et al., 2009) have revealed a high rate of genetic promiscuity due to extrapair (EP) copulations. Under natural conditions, 56% of the broods contained offspring from EP males, and approx. 21% of the nestlings were sired by EP father (Rosivall et al., 2009). Studies of other populations of this species have clearly shown the absence of intraspecific brood parasitism (Krist et al., 2005;Sheldon & Ellergren, 1996). Löhrl (1957) described that nest adoption by a male could occur at nests which the father had deserted, and he took an anecdotal note that once he observed a nest with one female and two males. In addition to the above patterns, we now provide detailed field and genetic data on the occurrence of one kind of cooperative breeding with three carers (two females and one male) and alloparental care in this species.

| MATERIAL S AND ME THODS
We collected data from a collared flycatcher population breeding in artificial nest box plots located in the Pilis-Visegrádi Mountains Nest box plots were checked at every 5 days in order to determine laying date, clutch size, and hatching date. During the breeding seasons of 2015 and 2016, in a brood-size manipulation experiment, we created trios (n = 10 in 2015, n = 9 in 2016) of certain nests and partially cross-fostered nestlings between two nests within trios (transferring 2 nestlings from nest A to nest B, and 4 nestlings from B to A; C was nonmanipulated), when the older nestlings of a nest were 2 days old. In our population, clutches generally contain asynchronously hatched nestlings: Usually, at least half of the nestlings hatch on the first day, and the others hatch on the next day. Rarely, 2-day asynchrony may occur between the oldest and the youngest nestlings, but these clutches could not be included into our brood-size manipulation experiment, according to our protocol.
We conducted video recordings to collect data on nestling feeding rates (number of visits with food delivery per hour) at the nests of the trios when the older chicks in each nest were 10 days old (n = 29 in 2015, n = 26 in 2016). As nestlings were of the same age within a trio, video recordings of nests of a trio were conducted on the same day and hour. We performed the video recordings between 0900 and 1830, mostly during the first half of the day. Earlier it had been described for this study population that no difference exists between morning and afternoon feeding rates of parents (Kiss et al., 2013). We avoided recording in midday hours (1200-1400).
Recording length was a few minutes more than 1 hr, and we analyzed Just after recording, we set up a spring trap at each nest box.
According to our general protocol, we finish trapping after we catch a female and a male bird (supposing that they are the only parents) or after a maximum of 1.5 hr. We ringed the adult birds if they were not ringed formerly. Blood samples (ca. 10-15 µl) were collected from the wing vein of nestlings and the captured parents for the purpose of paternity analysis. Blood samples were stored in absolute ethanol and kept at −20℃ until the analyses. DNA was extracted using an ammonium-acetate method (Nicholls et al., 2000). We assessed maternity and paternity by using ten polymorphic microsatellite loci (FhU2, FhU4, Fhy405, Fhy407, Fhy428, Fhy429, Fhy431, Fhy452, Cuµ4, Pdoµ5 (Ellegren, 1992;Primmer et al., 1996;Griffith et al., 1999;Gibbs et al., 1999;Leder et al., 2008)). All PCRs were run using the Type-it microsatellite PCR kit (QIAGEN) and the following thermal profile: 95℃ for 5 min, 30 cycles of 95℃ for 30 s, 56℃ for 30 s, 72℃ for 30 s, and a final step of 60℃ for 30 min.
PCR products were analyzed on ABI 3130xl Genetic Analyzer Maternity and paternity testing of the nestlings was performed manually by the exclusion method (Jones & Ardren, 2003). Assuming Mendelian inheritance, offspring were classified as not related to their putative parents if they did not match at one or more loci.

| RE SULTS
Breeding data indicated that the first egg in the nest box in question was laid on 27 April 2016. On 03 May, we found five eggs, which suggested that the last egg was laid on 01 May as females typically lay one egg on each consecutive day. On 07 May, the nest contained seven eggs. This means that there was a gap of a few days in egg laying, with a restart between 04 and 06 May. On May 20, we found six nestlings: three of them were 1 day old and the others were 2 days old, based on their appearances and body masses. We did not find any further propagule, indicating that one egg or hatchling had vanished. This missing offspring could not be sampled for DNA later.
The nest was subject to a cross-fostering experiment, and it was enlarged by three nestlings (from six to nine). Processing the video records, we observed that two females, which were distinguished by their different plumage patterns, and one male ( The genetic analysis of the six offspring that remained and the captured adults suggests that the captured female was related to none of the nestlings. On each locus, we could identify two alleles, either of which was present in every offspring. The most likely scenario is therefore that all nestlings shared the same mother (which we did not catch), even though we cannot completely exclude the existence of multiple mothers. The captured male sired only one progeny, and, assuming one mother, the other nestlings were sired by at least two other males. The captured female and male were unrelated or very distantly related to each other. The alleles of captured female, male, and offspring together with the assumed alleles of the putative mother are indicated in Table 1. Figure 2 shows a graphical representation of this unusual family setup.

| D ISCUSS I ON
In our case study, we detected cooperative parental care with two females and one male in the collared flycatcher, a species previously described as showing social monogamy (or polyterritorial polygyny) and biparental care. Neither of the two females nor the male that appeared on the video record seemed to be a prospector, because all of them showed active parental behavior in terms of providing nourishment and taking out nestling feces several times, and none of these are attributes of prospective behavior in this species (Doligez et al., 2004). Feeding rate of the male was similar to the mean feeding rate of the other males that raised enlarged broods.
In contrast, neither of the females showed the expected elevated feeding intensity. However, the two females together fed at a total rate (38) similar to the mean feeding rate of socially monogamous females of enlarged broods. It is clear that in spite of the presence of two females' assistance, the male's feeding intensity did not reduce, as well as from this triparental care the nestlings did not gain extra advantages at least with regard to their mean body mass at 12 days of age, close to fledging (12.4 g; in other enlarged broods: 12.7 g, Laczi et al., 2017).
The original nest contained eggs that had been laid in two separate turns. Based on this and the behavioral observations, one would conceive that the parenting mothers shared the genetic maternity too. If we take a look at the microsatellite patterns, we can draw quite a different conclusion, namely that there was probably only one genetic mother. This is also supported by the highly synchronous hatching (at most 1-day difference between the older and the F I G U R E 1 Cooperative parental care at a collared flycatcher nest. Each of the parents delivered food for the nestlings and took out the nestling feces younger nestlings) which suggested that the incubation did not start until after the laying of the fifth egg. The occurrence of laying gaps may be related to abiotic environmental stress conditions (Eeva & Lehikoinen, 2010;Nilsson & Svensson, 1993) or to social interactions (Low, 2008). Here, disturbances by predators may have led to the interruption in egg laying as there were nest predation events around this nest box during that breeding season (caused most probably by European pine marten (Martes martes) or beech marten (M. foina) based on the clues), aborting nesting in multiple nest boxes that had already contained eggs or nestlings.
The most likely cause of the appearance of the second, that is, the captured female (B), is also predation. Perhaps the inner drive of the female was so strong that after the depredation of her own nest, she continued feeding in a nearby nest box where hungry nestlings begged for food. In this case, the observed behavior was simply a sequel of reproductive error (Riedman, 1982) and did not provide any benefits. Alternatively, it might be beneficial in terms of using resources, or group membership may have enhanced antipredator defense (Emlen et al., 1991;Riedman, 1982). In any case, the lack of aggression between the two females (i.e., neither female chased away the other) suggests that the second female had already been parenting for a longer time and did not begin only on the day of the video recording.
With regard to the male, it has been shown in another collared flycatcher population that socially monogamous males adjust their provisioning rates according to their assumed share of paternity (Sheldon & Ellegren, 1998). We cannot completely exclude that the usually representative 1-hr sampling of parental care may have accidentally fallen on an unusual period at this particular nest. However, if the recorded care pattern is representative, we would speculate based on the feeding rate of the captured male that it was the genetic father of the majority of the original nestlings. By contrast, DNA analyses have shown us that most of the offspring were not sired by the captured male. In socially monogamous bird species, males that engage in extrapair copulations usually do not care for their offspring sired in other than their social mate's nest (Birkhead & Møller, 1992). In addition, when we estimated the height of the forehead patch of the male on the video record (using box entrance diameter as reference by Scanning Probe Image Processor, Image Metrology, Inc.), it was almost identical to the measured forehead patch size of the captured male (9.0 and 9.1 mm, respectively). Thus, we supposed that the captured male was identical to the one observed on the video record. If so, based on the genetic data, he must have been cuckolded by other males. In a socially monogamous system with extrapairs, it is not a curiosity if a male raises a nest with only one or even no own progeny (Gibbs et al., 1990;Kempenaers et al., 2001).
In conclusion, this brood was apparently cared for by a female that was related to none of the nestlings, the actual mother of the nestlings who had obtained extrapair fertilizations from at least two foreign males, and her heavily cuckolded social mate. Our finding is in accordance with the phenomenon that certain adverse environmental factors, such as elevated intensity of local predation, may dynamically TA B L E 1 Genetic profiles of the captured female, male, the nestlings, and the potential profile of the putative mother at ten microsatellite loci Note: The alleles were identified by capillary gel electrophoresis. Only one nestling was sired by the captured male (nestling 1; alleles that could be transmitted by the male are indicated in bold). At 3 loci, only one allele was present in every offspring; therefore, we could not identify the other alleles of these loci (marked with asterisks) for the putative mother.
rearrange the general mating and parental care scenario among a subset of individuals within a population. For example, in the socially monogamous blue tit (Cyanistes caeruleus), unbalanced sex ratio of reproductive birds, for example, the death of a neighboring parental male, may give rise to replacement polygyny (Kempenaers, 1994). In the same species, failed breeding could cause the type of nest adoption (Santema & Kempenaers, 2021) which has also been described in the collared flycatcher (Löhrl, 1957) (see Introduction). Mating and parental care relationships can be flexibly fine-tuned in accordance with availability of either sex not only in birds, but also in mammals, as described, for example, in the wild boar (Sus scrofa), where hunting on males reduced the degree of polygyny (Poteaux et al., 2009).
Our observations convey a warning message for investigations of reproductive behavioral patterns. After capturing a female and a male of a species that is expected to be socially monogamous and biparental, it may not be justified to automatically suppose that they are the mates of each other, and they are the parents or the only parents.
As another study on the pied flycatcher suggested, applying various methods together (i.e., genetic and behavioral) may be required to dig out the actual truth in some uncertain cases (Grinkov et al., 2018), thereby avoiding the misinterpretation caused by using one method only. Hence, a holistic approach can be highly beneficial when estimating actual mating preferences and realized reproductive success.

ACK N OWLED G M ENTS
This study was supported by NKFIH, grants number K124443 (GH), K120249 (BR), and the Pilis Park Forestry. The research was conducted in agreement with the current laws of Hungary with research permits from the regional nature conservation authority (KTVF 509-4/2012). We thank Pete Townshend for enlightening us that "the simple things you see are all complicated" (Substitute, 1978).

CO N FLI C T O F I NTE R E S T
The authors declare no competing interest.

F I G U R E 2
Relationships within an unusual collared flycatcher family with cooperative parental care. Video record revealed that three individuals (females (A) and (B), male (1)) were regularly delivering food and were taking out nestling feces. Microsatellite fingerprinting of the captured adults (marked with asterisks) and nestlings suggested that the captured female (B) was unrelated to the progenies, and probably all offspring shared one genetic mother, female (A). The captured male (1) was genetically related to only one offspring, represented by an egg with thick outline. The other nestlings were sired by at least two other males (2) and (3). Lines connect individuals with genetically shared offspring. Empty ellipse represents an egg that disappeared and was not sampled. For more details, see the text

DATA AVA I L A B I L I T Y S TAT E M E N T
All data are included in the publication.