Abstract

False feeding, where individuals refrain from delivering a food item to a begging dependent young, has been described in several cooperative bird and mammal species, but its function is still unclear. False feeding has been suggested to represent either a deceptive tactic of helpers aimed at showing off provisioning behavior to the rest of the group without paying the costs or a normal provisioning behavior of caregivers mediated by the trade-off between the hunger of the young and caregivers’ own conditions. Here, we employed an experimental approach to test whether false feeding in cooperatively breeding carrion crows responds plastically to variations of chicks’ and caregivers’ needs. In 4 different treatments, we manipulated the hunger of the brood and the conditions of group members by 1) experimentally feeding the chicks, 2) food-supplementing group members during the breeding season or 3) throughout the whole year, and 4) clipping 2 primary feathers from each wing of some individuals to increase the costs of flight. Breeders increased false feeding when the brood was food supplemented (treatment 1) and after their wings were clipped (treatment 4), whereas helpers did not change their false-feeding behavior in response to these treatments. Conversely, helpers decreased false feeding when food was supplemented year-round (treatment 3), whereas fed breeders did not show any significant difference compared with controls. These results indicate that a trade-off between chicks’ needs (current reproduction) and caregivers’ conditions (future reproduction) modulates the occurrence of false feeding, determining different responses in different group members.

False feeding, where individuals refrain from delivering a food item to a begging dependent young, has been documented in several cooperative bird and mammal species, such as Arabian babblers Turdoides squamiceps (Wright 1997), white-throated magpie-jays Calocitta formosa (Langen 1996), laughing kookaburras Dacelo novaeguineae (Legge 2000), bell miners Manorina melanophrys (Clarke 1984; Poiani 1993; McDonald et al. 2007), white-winged choughs Corcorax melanorhamphos (Boland et al. 1997), carrion crows Corvus corone corone (Canestrari et al. 2004), and meerkats Suricata suricatta (Clutton-Brock et al. 2005), but the function of this behavior is still unclear. In laughing kookaburras, which occasionally fail to deliver food to the nestlings (Legge 2000), false feeding has been interpreted as mistakes of inexperienced caregivers. A different explanation was suggested for white-winged choughs, where helpers personally consume the food brought to the nestlings on average in 11% of nest visits and preferentially do so when they are alone on the nest, away from potential onlookers (Boland et al. 1997). In this species, false feeders are sometimes punished by the other group members when detected, and the rates of false feeding drop in groups that are experimentally food-supplemented (Boland et al. 1997). False feeding in white-winged choughs seems therefore to represent a deceptive tactic aimed at satisfying helpers’ own hunger at the expenses of the brood and of the other adults (Boland et al. 1997; but see Clutton-Brock et al. 2005). Such deceptive tactics in cooperative species may indicate a signaling function of helping (Zahavi 1995), where individuals benefit from being observed provisioning the young but avoiding paying the costs (Putland 2001).

In meerkats, group members eat the prey item brought to a pup in 12% of cases (Clutton-Brock et al. 2005). Although helpers that contribute little to pup provisioning receive more aggression from dominants and show false feeding at higher rates compared with hardworking helpers, false feeding is unlikely to involve deception, as it can hardly be concealed and does not convey any obvious advantage to helpers. Authors suggest that meerkat helpers false feed when conflictive motivations arise (e.g., pups need and helpers’ own hunger) and individuals vacillate between alternative options (Clutton-Brock et al. 2005).

In bell miners, M. melanophrys, false feeds have been reported in several studies (Clarke 1984; Poiani 1993; McDonald et al. 2007). In 7.9% of nest visits, individuals carry no food or fail to deliver the prey load (totally or partially) to the nestlings (McDonald et al. 2007). Caregivers withhold the food when the brood is satiated, and they are more likely to deliver only part of the prey item when it contains a high proportion of lerp, which is relatively difficult to transfer to the nestlings (McDonald et al. 2007). False feeding does not decrease in the presence of onlookers and provokes no aggression when detected, excluding a deceptive function of this behavior. Alternatively, false feeding probably reflects normal provisioning behavior modulated by the adult's needs, chick's hunger, and prey type (McDonald et al. 2007).

Deception is similarly unlikely to account for the behavior of cooperative carrion crows, where individuals personally consume part or all the food brought to the nest in 9% of nest visits and breeders, especially females, false feed at the highest rates, independent of the presence of onlookers (Canestrari et al. 2004). In this species, like in meerkats and bell miners, false feeding has been suggested to occur as a result of a trade-off between the nestlings’ hunger, assessed by caregivers during nest visits, and the adult's own needs, which are assumed to be highest for breeding females who pay the highest costs of reproduction (Canestrari et al. 2004).

Although all previous studies have tested the “deceptive” function of false feeding, the idea that a trade-off between the needs of the young and the needs (or motivation) of caregivers plays a role in determining the occurrence of this behavior has not yet been tested experimentally (McDonald et al. 2007). In this paper, we present a comprehensive experimental approach with cooperatively breeding carrion crows, where we manipulated chicks’ hunger and group members’ needs in 4 different experimental treatments. In treatment one, we reduced chicks’ hunger by feeding them, and we compared the false-feeding behavior of group members confronted with satiated and hungry broods. In treatments 2 and 3, we reduced group members’ costs of provisioning by food supplementing some territories during the breeding season (treatment 2) and throughout the whole year (treatment 3), whereas other groups were not supplemented and used as control. We compared the proportion of false-feeding visits between members of fed and unfed groups. In treatment 4, we increased the costs of provisioning to members of different social groups. We did so by clipping 2 primary feathers from each wing of the experimental birds, which augmented the costs of flight, and we compared the proportion of false-feeding visits of the individuals before and after the treatment. If the balance between the chicks’ and group members’ needs is important in determining the occurrence of false feeding, we expect false feeding to augment when nestlings were fed (treatment 1) and/or when the costs of provisioning increased (treatment 4), and to decrease in group members living in food-supplemented territories (treatments 3 and 4).

MATERIALS AND METHODS

Study area and population

We have studied a population of carrion crows in a 45-km2 rural area in NW Spain (42°N, 5°W) since 1995. The study area represents a traditional Spanish low-intensity agricultural landscape, with a mosaic of crops, meadows, poplar and pine plantations, scrubs, oak forest patches, and uncultivated land. In this population, carrion crows form cohesive groups of up to 9 individuals through both delayed dispersal of offspring, which can remain on their natal territory with their parents for up to 4 years, and/or immigration of individuals that are related to the resident breeder of the same sex (Baglione, Marcos, and Canestrari 2002; Baglione et al. 2003). Sex ratio in the groups is skewed toward males, but females are found both among nondispersing offspring and immigrants. Unlike nondispersing offspring, adult immigrants (mainly males, but occasionally females, too) can share reproduction with the dominant pair (Baglione, Marcos, Canestrari, and Ekman 2002). Groups are therefore extended families, where the relatedness between group members and nestlings is high (Baglione et al. 2003). Both nondispersing offspring and immigrants can participate in nestling care (nest building, feeding the incubating female and the chicks, and nest sanitation), and nests can be attended by up to 5 caregivers (Canestrari et al. 2005).

The presence of helpers increases the total provisioning rate of the group and the annual fledgling production (Canestrari, Marcos, and Baglione 2008), but chick provisioning is costly to all caregivers, which lose weight in proportion to their effort (Canestrari et al. 2007). Breeders feed the chicks at higher rates compared with nonbreeding helpers, and among helpers males work harder than females, with no significant variation according to their relatedness to the nestlings (Canestrari et al. 2005). If food resources become temporarily abundant during the breeding season, all caregivers increase their body mass rather than augmenting chick provisioning rates (Canestrari et al. 2007). However, if food is abundant and predictable throughout the year, nonbreeding helpers, unlike breeders, increase their chick provisioning rates (Canestrari, Chiarati, et al. 2008). Furthermore, breeders reduce their own provisioning effort when assisted by more than one helper (Canestrari et al. 2007). These results indicate that in the trade-off between allocation of resources to the current brood and self-maintenance (i.e., future reproduction), “future” is important for crow group members and especially for breeders, which have higher probabilities of reproducing in the following breeding season (Canestrari, Chiarati, et al. 2008).

Classification of false feeds

In this work, we defined false feeds as events where caregivers withheld part or all the food brought to the nest, or took it back from the chick's mouth immediately after delivering, and consumed it themselves. It must be noted that, in our classification, false feeds did not include possible mistakes made by caregivers that carried a food item that was unsuitable or too big for the nestlings. In these cases, adults typically tried to deliver the prey to the chicks and retrieved it from the chick's gape one or several times, trying to break it. Eventually they swallowed the entire item or the part that was unpalatable for the chicks. These apparent mistakes occurred very rarely in crows and were never considered as false feeds. In the case of incomplete transfer of food load to the nestlings, we scored as false feeds only those events where a crow held the food in its crop and easily transferred part of it into the chick's gape, before swallowing the rest without even trying to break it or deliver it to the chicks. Similarly, we considered the action of retrieving food from a chick's gape as false feeding only when the recovered item did not look unsuitable for the nestlings. Observations on hand-raised crow chicks, fed with blended meat, revealed that nestlings sometimes do not swallow the item immediately but hold it at the bottom of the gape for up to 3 minutes. This suggests that caregivers sometimes have the opportunity to retrieve from a chick's gape suitable food items that are not yet swallowed. For each caregiver we calculated the “proportion of false-feeding visits” as the number of visits where false feeding occurred divided by the total number of visits where food was brought to the nest, and we transformed it with the arcsine square-root function for statistical analyses.

In a previous article on false feeding in crows (Canestrari et al. 2004), we described also other kinds of unusual behaviors at the nest, namely “empty visits,” where birds reach the nest carrying no food, and “stolen recoveries,” where a caregiver retrieves food from a chick's gape that was delivered by another individual. These behaviors are not considered in this present work because empty visits seem to result from to the need to carry out other activities at the nest, for example, sanitation (Canestrari et al. 2004), and stolen recoveries may be a form of kleptoparasitism rather than false feeds. An analysis of the function of the latter would have required separate analyses, which were impossible due the rarity of this behavior.

Bird capturing and banding

Each year since 1999, we have captured free-flying crows using 2-compartment walk-in traps and a remote-controlled snap trap (3 × 3 m) specifically developed for this species, which allowed catching 2–5 individuals at time (Baglione, Marcos, and Canestrari 2002). The individuals were banded with a unique combination of colored rings and plastic patagial wing tags (6.5 × 3.5 cm) and were aged as 1, 2, and older than 2 years according to the internal color of the upper mandible (Svensson 1992). The nestlings were banded just before they fledged (30 days after the first chick in the brood hatched). We collected between 50 μl and 200 μl of blood from the brachial vein of each banded individual for DNA extraction. Individual sex was determined with the P2/P8 molecular method (Griffiths et al. 1998) and then confirmed using primers 3007 and 3112 (Ellegren and Fridolfsson 1997).

Experimental treatments

Between 2003 and 2006, we studied 83 group-years where most members were banded and where individual breeding status could be assigned unambiguously because supernumerary birds were previous offspring of the breeding pair or immature immigrants (see Table 1 for a detailed summary of sample sizes). When one breeder was the only unbanded group member (n = 38), we inferred its sex based on the sex of the other banded breeder. Because only one of 44 breeders studied so far in this population was younger than 3 years (Canestrari D, unpublished data), these unbanded breeders were aged as adults. Eleven groups contained 2–3 unbanded individuals that were excluded from the analyses because their breeding status and sex could not be determined. Throughout the study years, we carried out a total of 4 different experimental treatments in order to manipulate the needs of nestlings and group members, and we video-recorded activity at the nests by placing camouflaged microvideo cameras 2.5 m away from the nests (Canestrari et al. 2005). The video cameras were set up at least 5 days before the start of video-recorded observations and were removed after the chicks had fledged. The video recorder and the batteries powering the whole system were placed at the bottom of the nest tree, and therefore the routine change of batteries and videotapes provoked little disturbance to the nest. In all treatments, video-recorded observations were carried out during a time span ranging between 7:30 and 13:30.

Table 1

Description of sample sizes

Experimental treatment (procedure) Number of groups (experimental/control) Number of breeders (sex) Number of nonbreeders (sex) Number of video-recorded observation bouts per nest (number of hours/bout) 
Food supplementation to chicks (experimental and control treatments to the same individuals) 10 (experimental and control) 10 ♂ 8 ♂ 1 (4) experimental 
10 ♀ 3 ♀ 1 (4) control 
Food supplementation to group members during breeding season (experimental and control treatments to different groups) 12 experimental 12 ♂ 10 ♂ 3–5 (4) 
11 ♀ 10 ♀ 
14 control 16 ♂ 18 ♂ 3–5 (4) 
13 ♀ 10 ♀ 
Food supplementation to group members year-round (experimental and control treatments to different groups) 7 experimental 6 ♂ 5 ♂ 3–5 (4) 
7 ♀ 5 ♀ 
14 control 11 ♂ 14 ♂ 3–5 (4) 
9 ♀ 6 ♀ 
Wing clipping to group members (collection of data from the same individual before and after the treatment) 17 (experimental and control) 8 ♂ 7 ♂ 3 (4) experimental 
1 ♀ 1 ♀ 3 (4) control 
Experimental treatment (procedure) Number of groups (experimental/control) Number of breeders (sex) Number of nonbreeders (sex) Number of video-recorded observation bouts per nest (number of hours/bout) 
Food supplementation to chicks (experimental and control treatments to the same individuals) 10 (experimental and control) 10 ♂ 8 ♂ 1 (4) experimental 
10 ♀ 3 ♀ 1 (4) control 
Food supplementation to group members during breeding season (experimental and control treatments to different groups) 12 experimental 12 ♂ 10 ♂ 3–5 (4) 
11 ♀ 10 ♀ 
14 control 16 ♂ 18 ♂ 3–5 (4) 
13 ♀ 10 ♀ 
Food supplementation to group members year-round (experimental and control treatments to different groups) 7 experimental 6 ♂ 5 ♂ 3–5 (4) 
7 ♀ 5 ♀ 
14 control 11 ♂ 14 ♂ 3–5 (4) 
9 ♀ 6 ♀ 
Wing clipping to group members (collection of data from the same individual before and after the treatment) 17 (experimental and control) 8 ♂ 7 ♂ 3 (4) experimental 
1 ♀ 1 ♀ 3 (4) control 

Treatment 1

Food supplementation to nestlings. In 2006, we chose 10 family groups containing a breeding pair and 1–2 nondispersing offspring (Table 1) and assigned their nestlings (n = 38) to a food supplementation experiment. The food used was a canned dog food specially formulated for pups (brand Hill's), rich in proteins and highly palatable for crow nestlings. When chicks were 15 days old, they were removed on 2 consecutive days from the nest by one member of the research team and were transferred to the bottom of the tree in a cotton bag. On the “experimental” day, the chicks were fed by a second researcher until they stopped begging even when stimulated by touching their beaks with the fingers, and on the “control” day they were given 0.5 ml of drinking water using a plastic syringe with the needle removed. Such a small amount of water did not satiate the chicks, which continued begging after ingesting the liquid (see RESULTS). The order of the 2 treatments was randomized among nests. The whole manipulation (including tree climbing, removing the chicks from the nest, performing experimental or control treatment, and returning the chicks into the nest) lasted not more than 35 min. Immediately afterward, we carried out a video-recorded nest observation of 4 h after each treatment. From the video recordings, we measured the begging intensity of individual chicks, for each nest visit made by caregivers, by scoring nestlings’ body posture from 0 (no begging) to 4 (intense begging) on the caregiver's arrival. A chick's begging intensity was scored as 0 if the chick laid in the nest without moving its head on the caregiver's arrival, 1 if it raised only its head and gaped, 2 if it gaped raising its head with the neck totally stretched, 3 if it gaped raising the whole body, and 4 if it begged raising the whole body and hanging its wings (Kilner 1995). For each nest, we calculated the average begging intensity of the brood in each treatment (experimental and control).

Treatment 2

Food supplementation to group members during the breeding season. At the beginning of breeding seasons 2003 and 2004, we paired groups that were similar in size and composition and randomly assigned one group of each pair to the experimental treatment. We fed the experimental groups with canned dog food every day in the middle of the territory (about 200 m away from the nest) from the beginning of the breeding season (after the first nest was completed but just before egg laying, which typically occurs 6–9 days after the nest is built; Marcos JM, unpublished data) until the chicks fledged or the nest failed at egg or nestling stage. Video recording sessions in all experimental groups confirmed that the target crows were actually taking the food and that all group members had access to it. Typically, the supplementary food was removed and eaten or stored within 20 min of our departure. The dog food used (brand DIA, chicken and beef flavor) is very palatable for crows, and this commercial mixture of meat and vegetables (78.8% and 21.2%, respectively) falls within the natural range of variability of crow's diet in terms of proportions of animal and vegetal components (respectively from 82.6–17.4% to 41.7–58.3%; Cramp and Perrins 1994). The individual daily amount of food supplemented (about 200 grams/crow/day, corresponding to ca. 1071 Kj/crow/day) was calculated in order to provide an energy intake likely to affect the cost of provisioning (Nagy et al. 1999) without exceeding the range of natural variation of food resources in crow territories during the breeding season (Canestrari et al. 2007). Due to early nest failures, some of the chosen groups could not be sampled. Eventually, we obtained data from 14 unfed groups and 12 fed groups (Table 1). Although this disrupted our original matching of control and experimental territories according to group size (see above), the final sample was still balanced, with no significant difference in group size between fed and unfed territories (average group size ± standard error [SE] fed groups = 3.25 ± 0.22; unfed groups = 3.33 ± 0.19; t-test = 0.298; degrees of freedom [df] = 25; P = 0.8). For each nest, we collected 3–5 recording bouts of 4 h between day 10 and day 15 from hatching. For further details on the experimental treatment see Canestrari et al. 2007.

Treatment 3

Food supplementation to group members throughout the whole year. At the end of breeding seasons 2003 and 2004, we paired groups of similar size and randomly assigned one group of each pair to the experimental treatment (43 groups in total). We fed the experimental groups with 400 g of canned dog food (see above, Canestrari et al. 2007 and Canestrari, Chiarati, et al. 2008 for details on the brand and the calculation of the daily amount of food supplemented) and 200 g of maize per group member 3 times a week in the middle of the territory (about 200 m away from the nest) until the end of the following breeding season (i.e., until the chicks fledged or the nest failed at egg or nestling stage). During breeding seasons 2004 and 2005, we collected 3–5 video-recorded observations at the nests of 4 h each, between day 10 and day 15 from hatching. Due to early nest failures or inaccessibility of nests for recording, not all the initial study groups could be sampled. Ultimately, we collected information on 7 fed groups and 14 unfed groups (Table 1). Although this disrupted our original matching of control and experimental territories according to group size, the final sample was still balanced, with no significant difference in group size between fed and unfed territories (average group size ± SE fed groups = 4.57 ± 0.49; unfed groups = 4.29 ± 0.28; t-test = 0.38; df = 19; P = 0.7). Different groups were studied in the 2 different years. However, in 2004, this experiment overlapped with the experiment described in treatment 2 which was carried out in 2003 and 2004. In 2004, 9 control territories (unfed) were shared between the 2 experiments.

Treatment 4

Wing clipping of group members. At the beginning of breeding seasons 2006 and 2007 we chose a total of 17 groups containing 33 recognizable breeders (17 males and 16 females), 27 nondispersing offspring (17 males and 10 females), and a 2-year-old male immigrant that was classified as nonbreeder due to his young age. When chicks were 10–13 days old, we made the first set of 3 video-recorded observations (one per day, each 4 h) at the nests. We immediately analyzed the video recordings in order to determine which group members were actually provisioning the chicks. Between day 14 and day 16, we captured one provisioning group member for each group and experimentally reduced its body condition by clipping close to the base the 7th and 9th primary feathers of both wings. The removal of those feathers, the area of which averaged 41.5 cm2 per wing, reduced total wing area (average 399 cm2) by approximately 10.4%. The wing loading of a 500 g crow would increase by 11.6% (12.28 N m−2 compared with 13.71 N m−2) (Pennycuick 1989). We eventually clipped 9 breeders and 8 nonbreeders (Table 1). Starting from the day after the treatment, we carried out a total of 3 video-recorded sessions of 4 h per nest, one session per day. Each individual was therefore sampled before and after the treatment.

Statistical analyses

Statistical analyses were performed using Genstat 10.0. Unless stated otherwise, all data were analyzed with linear mixed models (LMM). Year and territory identity, which pooled data from individuals of the same group, were fitted as random factors. Where appropriate, group member identity was fitted as a second random factor to account for repeated measures of the same individual under different experimental conditions. Potential explanatory variables that gave nonsignificant results (P > 0.1) were sequentially removed until the model only included those terms for which elimination would have significantly reduced the explanatory power (minimal model). Significant probability values were derived from having all significant terms fitted in the model together, whereas those of nonsignificant terms were obtained from having all significant terms in the model and each nonsignificant term fitted individually (Crawley 2002; Russell et al. 2003). In the results, values for nonsignificant interactions are omitted. Tables show average effects ± SE of significant factors.

Treatment 1

In order to determine whether our food supplementation treatment affected chicks’ needs, we used the average begging score as the dependent variable, and we ran a Wilcoxon matched-pairs test to compare the begging intensity of the same brood in experimental and control treatments. To analyze the effect of food supplementation to chicks on the caregivers’ proportion of false-feeding visits, we fitted into a LMM the following explanatory variables: breeding status (breeder/helper), sex, brood size, number of caregivers, experimental treatment (food/water), experimental treatment × breeding status, and experimental treatment × sex. Territory and caregiver identity were fitted as random factors.

Treatments 2 and 3

We analyzed the effect of food supplementation to group members (both during the breeding season and year-round; treatments 2 and 3, respectively) on the proportion of false feeding visits by fitting into a LMM the following explanatory variables: breeding status (breeder/helper), sex, brood size, number of caregivers, experimental treatment (fed/unfed), experimental treatment × breeding status, and experimental treatment × sex. Year and territory identity were fitted as random factors.

Treatment 4

We tested the effect of wing clipping of group members on their proportion of false-feeding visits by comparing the behavior of the experimental individuals before and after the treatment. We fitted into a LMM the following explanatory variables: breeding status (breeder/helper), brood size, number of caregivers, experimental treatment (before/after the treatment), and experimental treatment × breeding status. Sex was omitted as explanatory factor due to the low number of females (n = 2) in the sample. Year and caregiver identity were fitted as random factors. In this experiment, individual changes in false-feeding behavior after the treatment could be potentially confounded by an effect of time, as control and experimental observations were necessarily made in a set order. However, if this was the case, we should observe a change in the proportion of false-feeding visits over successive observation bouts among untreated individuals too. For 20 control individuals chosen at random from experiments 1, 2, and 3, we calculated the proportion of false-feeding visits for each hour of observation (n = 235 data points). With a LMM, we tested whether the proportion of false-feeding visits changed over successive observation bouts or within each bout. Individual identity and territory identity were fitted as random factors, whereas sex, group size, brood size, number of observation bout, and hour of observation within each bout were fitted as explanatory variables. The proportion of false-feeding visits did not change over successive observation bouts (Wald statistic = 1.16; df = 1,193.4; P = 0.28), suggesting that any observed response to the wing-clipping treatment could not be confounded by the effect of time. In addition, the lack of variation in the proportion of false-feeding visits within each bout (Wald statistic = 0.29; df = 1,221.9; P = 0.59) indicated that false-feeding behavior was not influenced by the possible disturbance caused by the presence of the video cameras or the observer that started the recording system at the bottom of the nest tree. If this was the case, the proportion of false-feeding visits should have decreased over successive observation bouts and/or within each observation bout as a consequence of habituation, but this did not occur.

Ethical note

All bird manipulations described here were authorized by Junta de Castilla y León (permit EP/177/2003) and had no visible adverse effect. No crows were injured nor did any abandon the nest or territory as a consequence of capture. The model of patagial wing tags used has been successfully employed in several corvid populations with no adverse effect on survivorship or social relationships (Caffrey 2000; Canestrari et al. 2007). The birds never showed any visible reaction to the video cameras, and video-recorded nests did not show a higher rate of brood failure compared with control nonvideo-recorded nests (Canestrari et al. 2005). Wing-clipped individuals did not abandon the territory or the nest. One year after the treatment, all but one clipped individual were observed alive in their territories. Those birds that were clipped as breeders reproduced normally the following year. An alternative method to increase costs of flying is to attach weights to the bird's legs (Pennycuick 1989). However, weights must be removed at the end of the experiment in order to avoid permanent effects (Cuthill 1991) and catching the same crow twice is extremely difficult. This is why we decided to remove 2 primary feathers, which represented a temporary handicap resembling the natural situation of molt (Cuthill 1991; Velando and Alonso-Alvarez 2003).

RESULTS

Experimental food supplementation to the chicks (treatment 1) significantly decreased nestlings’ begging intensity (without suppressing it) during the 4 h after the manipulation, compared with the control treatment applied to the same sample on a different day (median begging score fed chicks = 1.43; unfed chicks = 3.35; Wilcoxon matched-pairs test Z = 2.37, P = 0.018). Breeders increased their proportion of false-feeding visits when chicks were experimentally fed while helpers did not, as indicated by the significant effect of the interaction between breeding status and experimental treatment (Table 2, Figure 1).

Table 2

Effect of treatment 1 (food supplementation to the chicks) on the proportion of false-feeding visits

Model terms Wald statistic df P Random term estimated variance component ± SE Average effect ± SE 
Territory identitya    −0.02 ± 0.012  
Caregiver identitya    0.07 ± 0.04  
Breeding status 0.02 1,29 0.89   
Sex 0.06 1,28 0.81   
Age 0.18 2,20.3 0.67   
Brood size 5.86 1,1.5 0.25   
Number of caregivers 0.04 1,9.5 0.85   
Experimental treatment (food/water) 10.02 1,29 0.004   
Breeding status × experimental treatment (food/water) 5.62 1,29 0.025   
Model terms Wald statistic df P Random term estimated variance component ± SE Average effect ± SE 
Territory identitya    −0.02 ± 0.012  
Caregiver identitya    0.07 ± 0.04  
Breeding status 0.02 1,29 0.89   
Sex 0.06 1,28 0.81   
Age 0.18 2,20.3 0.67   
Brood size 5.86 1,1.5 0.25   
Number of caregivers 0.04 1,9.5 0.85   
Experimental treatment (food/water) 10.02 1,29 0.004   
Breeding status × experimental treatment (food/water) 5.62 1,29 0.025   
a

Random factors.

Figure 1

Response to treatment 1. Proportion of false-feeding visits by breeders and helpers to fed and unfed chicks. Sample sizes are given above bars.

Figure 1

Response to treatment 1. Proportion of false-feeding visits by breeders and helpers to fed and unfed chicks. Sample sizes are given above bars.

When group members were supplemented with food during the breeding season (treatment 2) they did not significantly change the proportion of false-feeding visits (Table 3, Figure 2a). Generally, breeders showed a higher proportion of false-feeding visits than helpers, whereas sex, age, and brood size had no significant effect (Table 3). When food was supplemented throughout the whole year (treatment 3) fed helpers significantly decreased the proportion of false-feeding visits compared with unfed ones, whereas experimental and control breeders false fed at a comparable rate, as shown by the significant interaction between breeding status and experimental treatment (Table 4, Figure 2b). Conversely, experimental wing clipping of caregivers (treatment 4) determined an increase in the proportion of false-feeding visits among breeders but not among helpers, as indicated by the significant effect of the interaction between experimental treatment and breeding status (Table 5, Figure 3).

Table 3

Effect of treatment 2 (food supplementation to adults during the breeding season) on the proportion of false-feeding visits

Model terms Wald statistic df P Random term estimated variance component ± SE Average effect ± SE 
Territory identitya    0.003 ± 0.0027  
Yeara    −0.0004 ± 0.0003  
Breeding status 8.32 1,56.7 0.006   
Sex 0.04 1,55 0.84   
Age 0.35 1,73.2 0.55   
Brood size 0.07 1,38.2 0.79   
Number of caregivers 3.35 1,25.4 0.03   
Experimental treatment (fed/unfed) 0.22 1,32.5 0.64   
Minimal model      
Constant     0.18 ± 0.02 
Breeding status      
Breeders     
Helpers     −0.018 ± 0.02 
Number of caregivers     0.047 ± 0.02 
Model terms Wald statistic df P Random term estimated variance component ± SE Average effect ± SE 
Territory identitya    0.003 ± 0.0027  
Yeara    −0.0004 ± 0.0003  
Breeding status 8.32 1,56.7 0.006   
Sex 0.04 1,55 0.84   
Age 0.35 1,73.2 0.55   
Brood size 0.07 1,38.2 0.79   
Number of caregivers 3.35 1,25.4 0.03   
Experimental treatment (fed/unfed) 0.22 1,32.5 0.64   
Minimal model      
Constant     0.18 ± 0.02 
Breeding status      
Breeders     
Helpers     −0.018 ± 0.02 
Number of caregivers     0.047 ± 0.02 
a

Random factors.

Table 4

Effect of treatment 3 (food supplementation to adults year-round) on the proportion of false-feeding visits

Model terms Wald statistic df P Random term estimated variance component ± SE Average effect ± SE 
Territory identitya    0.002 ± 0.007  
Yeara    0.006 ± 0.011  
Breeding status 1.07 1,64.6 0.31   
Sex 0.22 1,51.5 0.64   
Age 0.3 1,64.9 0.58   
Brood size 0.16 1,20.3 0.69   
Number of caregivers 1.65 1,21.7 9.21   
Experimental treatment (fed/unfed) 0.97 1,8.3 0.35   
Breeding status × experimental treatment 4.41 1,56.5 0.04   
Model terms Wald statistic df P Random term estimated variance component ± SE Average effect ± SE 
Territory identitya    0.002 ± 0.007  
Yeara    0.006 ± 0.011  
Breeding status 1.07 1,64.6 0.31   
Sex 0.22 1,51.5 0.64   
Age 0.3 1,64.9 0.58   
Brood size 0.16 1,20.3 0.69   
Number of caregivers 1.65 1,21.7 9.21   
Experimental treatment (fed/unfed) 0.97 1,8.3 0.35   
Breeding status × experimental treatment 4.41 1,56.5 0.04   
a

Random factors.

Table 5

Effect of treatment 4 (wing clipping to adults) on the proportion of false-feeding visits

Model terms Wald statistic df P Random term estimated variance component ± SE Average effect ± SE 
Caregiver identitya    0.01 ± 0.011  
Yeara    0.02 ± 0.03  
Breeding status 4.04 1,14 0.064   
Age 0.34 1,14 0.56   
Brood size 0.23 1,14.3 0.63   
Number of caregivers 1.09 1,14.3 0.29   
Experimental treatment (before/after clipping) 4.17 1,15 0.059   
Breeding status × experimental treatment (before/after clipping) 4.44 1,15 0.05   
Minimal model      
Constant     0.26 ± 0.12 
Breeding status      
Breeders     
Helpers     −0.03 ± 0.09 
Experimental treatment      
Before clipping     
After clipping     0.24 ± 0.08 
Model terms Wald statistic df P Random term estimated variance component ± SE Average effect ± SE 
Caregiver identitya    0.01 ± 0.011  
Yeara    0.02 ± 0.03  
Breeding status 4.04 1,14 0.064   
Age 0.34 1,14 0.56   
Brood size 0.23 1,14.3 0.63   
Number of caregivers 1.09 1,14.3 0.29   
Experimental treatment (before/after clipping) 4.17 1,15 0.059   
Breeding status × experimental treatment (before/after clipping) 4.44 1,15 0.05   
Minimal model      
Constant     0.26 ± 0.12 
Breeding status      
Breeders     
Helpers     −0.03 ± 0.09 
Experimental treatment      
Before clipping     
After clipping     0.24 ± 0.08 
a

Random factors.

Figure 2

Response to treatments 2 (a) and 3 (b). (a) Proportion of false-feeding visits by breeders and helpers of fed and unfed territories that were treated during the breeding season. Sample sizes are given above bars; (b) Proportion of false-feeding visits by breeders and helpers of fed and unfed territories that were treated throughout the whole year. Sample sizes are given above bars.

Figure 2

Response to treatments 2 (a) and 3 (b). (a) Proportion of false-feeding visits by breeders and helpers of fed and unfed territories that were treated during the breeding season. Sample sizes are given above bars; (b) Proportion of false-feeding visits by breeders and helpers of fed and unfed territories that were treated throughout the whole year. Sample sizes are given above bars.

Figure 3

Response to treatment 4. Proportion of false-feeding visits by experimental breeders and helpers before and after clipping. Sample sizes are given above bars.

Figure 3

Response to treatment 4. Proportion of false-feeding visits by experimental breeders and helpers before and after clipping. Sample sizes are given above bars.

DISCUSSION

Our data demonstrated that false feeding in carrion crows responded to variation in both chicks’ hunger and caregivers’ needs. This indicates that crow caregivers which personally consume the food brought to the nest do not merely commit “mistakes” (Legge 2000), which are expected to occur independently of chicks’ and adults’ condition. Instead, the response of false-feeding behavior to our experimental manipulations revealed a fine adjustment of provisioning effort by crow caregivers, suggesting that decisions on investment of resources did not only involve changes in the frequency of nest visits but also the decision on how to invest every single food item brought to the young.

When chicks were fed and decreased begging intensity (treatment 1) and when chick provisioning became more costly for wing-clipped individuals (treatment 4) breeders increased their proportion of false-feeding visits, whereas helpers did not respond to the treatments by changing their false-feeding behavior. Conversely, when food became abundant year-round (treatment 3), fed helpers decreased the proportion of false-feeding visits compared with unfed ones, whereas experimental and control breeders false fed at similar rates. These results mirror previous findings on chick provisioning strategies that showed that, in the trade-off between investment in self-maintenance and current brood, breeders are more likely to invest in themselves (i.e., future reproduction), when conditions change, whereas helpers are more likely to increase investment in the current brood when the quality of their territories improves year-round (Canestrari et al. 2007; Canestrari, Chiarati, et al. 2008). Because crow breeders are generally more likely to reproduce in the following year than helpers (Baglione et al. 2005; Canestrari, Chiarati, et al. 2008), provisioning decisions in this population support the idea that individuals with higher possibilities to reproduce in the future show relatively greater investment in self-maintenance (Velando and Alonso-Alvarez 2003). The fine correspondence between the patterns of food provided (Canestrari et al. 2007 and Canestrari, Chiarati, et al. 2008) and food withheld (this study) boosts the conclusion that false feeding is a true provisioning strategy in carrion crows.

When additional food was provided to the territories during the breeding season (treatment 2), neither breeders nor helpers significantly decreased their proportion of false-feeding visits. This result differs from that found in white-winged choughs and meerkats, where adults false fed less when supplemented with food (Boland et al. 1997; Clutton-Brock et al. 2005). However, this makes sense in the case of cooperative carrion crows. Previous studies showed that crow group members did not increase their provisioning rates when they were fed during the breeding season, preferring to allocate these additional resources in self-maintenance by reducing their body mass loss (Canestrari et al. 2007). The fact that they did not reduce false feeding, and therefore did not invest more in the current brood, when provided with additional food is consistent with this result. On the other hand, when crows were food supplemented year-round, helpers (unlike breeders) increased their provisioning effort (Canestrari, Chiarati, et al. 2008) and false fed less (see above).

So far, false feeding has been described only in cooperative social systems. This is surprising because the costs of parental care and the trade-off between current and future reproduction must be weighed carefully in most animal species (Alonso-Alvarez and Tella 2001; Velando and Alonso-Alvarez 2003). We may therefore expect both biparental and cooperative species to show false-feeding behavior. One simple explanation is that this behavior has been overlooked in many species and that it may be more frequent than expected. Alternatively, false feeding may occur more frequently in cooperative species because the presence of several caregivers can better compensate for a reduction in provisioning rate by some group member (Hatchwell and Davies 1990; Crick 1992). Interestingly, the proportion of false-feeding visits in cooperative crows also increased with the number of provisioning group members, supporting the idea that false feeding may be afforded only by species with compensatory helpers. In addition, the presence of several caregivers in cooperative species may determine an overall higher provisioning frequency to the brood, which decreases chick begging intensity and ultimately determines the occurrence of false feeds. Indeed, in cooperative crows, the total provisioning rate of a group increases with the number of caregivers (Canestrari, Marcos and Baglione 2008) and may explain the higher frequency of false feeds in larger groups (this study).

In conclusion, our study experimentally supported the idea that false feeding is part of the provisioning behavior repertoire of carrion crows that is influenced by the relative conditions of chicks and caregivers (Canestrari et al. 2004; Clutton-Brock et al. 2005; McDonald et al. 2007) and by the trade-off between current and future reproduction. Further investigation is needed to clarify whether false feeding prevails in cooperative species with respect to biparental ones.

FUNDING

This work was supported by the Spanish Ministry of Science and Innovation (project grants CGL2005-02083/BOS, CGL2008-01829BOS to V.B. and “Juan de la Cierva” program-FSE to D.C.), the program COCOR ‘Cooperation in Corvids’, which forms part of the ESF-EUROCORES program TECT ‘The Evolution of Cooperation and Trading’ (SEJ2007-29836-E), the “Ramón y Cajal” program (FEDER “Fondo Europeo de Desarrollo Regional”, FSE “Fondo Social Europeo”) to V.B., the Gates Cambridge Trust scholarship (to D.C.), Emmanuel College fieldwork grant (to D.C.), and the Xunta de Galicia (Isidro Parga Pondal program to M.V.).

We are grateful to Rebecca Heiss, Lizzie Rowe, and Nicholas Tomasino for helpful comments on the manuscript, and to Rudy Valfiorito and Gloria Robles for help in the field. All experiments comply with the law of the country in which they were performed, and all bird manipulations were authorized by Junta de Castilla y León.

References

Alonso-Alvarez
C
Tella
JL
Effect of experimental food restriction and body-mass changes on avian T-cell mediated immune response
Can J Zool
 , 
2001
, vol. 
79
 (pg. 
101
-
105
)
Baglione
V
Canestrari
D
Marcos
JM
Ekman
J
Kin selection in cooperative alliances of carrion crows
Science
 , 
2003
, vol. 
300
 (pg. 
1947
-
1949
)
Baglione
V
Marcos
JM
Canestrari
D
Cooperatively breeding groups of Carrion Crow (Corvus corone corone) in northern Spain
Auk
 , 
2002
, vol. 
119
 (pg. 
790
-
799
)
Baglione
V
Marcos
JM
Canestrari
D
Ekman
J
Direct fitness benefits of group living in a complex cooperative society of carrion crows, Corvus corone corone
Anim Behav
 , 
2002
, vol. 
64
 (pg. 
887
-
893
)
Baglione
V
Marcos
JM
Canestrari
D
Griesser
M
Andreotti
G
Bardini
C
Bogliani
G
Does year-round territoriality rather than habitat saturation explain delayed natal dispersal and cooperative breeding in the carrion crow?
J Anim Ecol
 , 
2005
, vol. 
74
 (pg. 
842
-
851
)
Boland
CRJ
Heinsohn
R
Cockburn
A
Deception by helpers in cooperatively breeding white-winged choughs and its experimental manipulation
Behav Ecol Sociobiol
 , 
1997
, vol. 
41
 (pg. 
251
-
256
)
Caffrey
C
Marking crows
N Am Bird Band
 , 
2000
, vol. 
26
 (pg. 
146
-
148
)
Canestrari
D
Chiarati
E
Marcos
JM
Ekman
J
Baglione
V
Helpers but not breeders adjust provisioning effort to year-round territory resource availability in carrion crows
Anim Behav
 , 
2008
, vol. 
76
 (pg. 
943
-
949
)
Canestrari
D
Marcos
JM
Baglione
V
False feedings at the nests of carrion crows Corvus corone corone
Behav Ecol Sociobiol
 , 
2004
, vol. 
55
 (pg. 
477
-
483
)
Canestrari
D
Marcos
JM
Baglione
V
Effect of parentage and relatedness on the individual contribution to cooperative chick care in carrion crows
Behav Ecol Sociobiol
 , 
2005
, vol. 
52
 (pg. 
422
-
428
)
Canestrari
D
Marcos
JM
Baglione
V
Costs of chick provisioning in cooperatively breeding carrion crows: an experimental study
Anim Behav
 , 
2007
, vol. 
73
 (pg. 
349
-
357
)
Canestrari
D
Marcos
JM
Baglione
V
Reproductive success increases with group size in cooperative carrion crows Corvus corone corone
Anim Behav
 , 
2008
, vol. 
75
 (pg. 
403
-
416
)
Clarke
MF
Co-operative breeding by the Australian Bell Miner Manorina melanophrys Latham: a test of kin selection theory
Behav Ecol Sociobiol
 , 
1984
, vol. 
14
 (pg. 
137
-
146
)
Clutton-Brock
TH
Russell
AF
Sharpe
LL
Jordan
NR
False feeding” and aggression in meerkat societies
Anim Behav
 , 
2005
, vol. 
69
 (pg. 
1273
-
1284
)
Cramp
S
Perrins
CM
The birds of the Western Paleartic
1994
, vol. 
Vol. 8
 
Crows to finches. Oxford
Oxford University Press
Crawley
MJ
Statistical computing
 , 
2002
London
J. Wiley
Crick
HQP
Load-lightening in cooperatively breeding birds and the costs of reproduction
Ibis
 , 
1992
, vol. 
134
 (pg. 
56
-
61
)
Cuthill
I
Field experiments in animal behavior—methods and ethics
Anim Behav
 , 
1991
, vol. 
42
 (pg. 
1007
-
1014
)
Ellegren
H
Fridolfsson
AK
Male-driven evolution of DNA sequences in birds
Nat Genet
 , 
1997
, vol. 
17
 (pg. 
182
-
184
)
Griffiths
R
Double
MC
Orr
K
Dawson
RJG
A DNA test to sex most birds
Mol Ecol
 , 
1998
, vol. 
7
 (pg. 
1071
-
1075
)
Hatchwell
BJ
Davies
NB
Provisioning of nestlings by dunnocks, Prunella Modularis, in pairs and trios—compensation reactions by males and females
Behav Ecol Sociobiol
 , 
1990
, vol. 
27
 (pg. 
199
-
209
)
Kilner
R
When do canary parents respond to nestling signals of need
Proc R Soc Lond Ser B Biol Sci
 , 
1995
, vol. 
260
 (pg. 
343
-
348
)
Langen
TA
The mating system of the white-throated magpie-jay Calocitta formosa and Greenwood's hypothesis for sex-biased dispersal
Ibis
 , 
1996
, vol. 
138
 (pg. 
506
-
513
)
Legge
S
Helper contributions in the cooperatively breeding laughing kookaburra: feeding young is no laughing matter
Anim Behav
 , 
2000
, vol. 
59
 (pg. 
1009
-
1018
)
McDonald
PG
Kazem
AJN
Wright
J
A critical analysis of ‘false-feeding’ behavior in a cooperatively breeding bird: disturbance effects, satiated nestlings or deception?
Behav Ecol Sociobiol
 , 
2007
, vol. 
61
 (pg. 
1623
-
1635
)
Nagy
KA
Girard
IA
Brown
TK
Energetics of free-ranging mammals, reptiles and birds
Annu Rev Nutr
 , 
1999
, vol. 
19
 (pg. 
247
-
277
)
Pennycuick
CJ
Bird flight performance: a practical calculation manual
 , 
1989
Oxford
Oxford University Press
Poiani
A
Social structure and the development of helping behaviour in the bell miner (Manorina melanophris, meliphagidae)
Ethology
 , 
1993
, vol. 
93
 (pg. 
62
-
80
)
Putland
D
Has sexual selection been overlooked in the study of avian helping behaviour?
Anim Behav
 , 
2001
, vol. 
62
 (pg. 
811
-
814
)
Russell
AF
Brotherton
PNM
McIlrath
GM
Sharpe
LL
Clutton-Brock
TH
Breeding success in cooperative meerkats: effects of helper number and maternal state
Behav Ecol
 , 
2003
, vol. 
14
 (pg. 
486
-
492
)
Svensson
L
Identification guide to European Passerines
 , 
1992
4th ed
Stockholm (Sweden)
British Trust for Ornithology
 
Privately Published
Velando
A
Alonso-Alvarez
C
Differential body condition regulation by males and females in response to experimental manipulations of brood size and parental effort in the blue-footed booby
J Anim Ecol
 , 
2003
, vol. 
72
 (pg. 
846
-
856
)
Wright
J
Helping-at-the-nest in Arabian babblers: signalling social status or sensible investment in chicks?
Anim Behav
 , 
1997
, vol. 
54
 (pg. 
1439
-
1448
)
Zahavi
A
Altruism as a handicap—the limitations of kin selection and reciprocity
J Avian Biol
 , 
1995
, vol. 
26
 (pg. 
1
-
3
)