Behavioral Ecology Vol. 11 No. 1: 110-114
© 2000 International Society for Behavioral Ecology
Social context affects risk taking by a satellite species in a mixed-species foraging group
Behavioral Ecology Group, Department of Evolution, Ecology, and Organismal Biology, The Ohio State University, 1735 Neil Avenue, Columbus, OH 43210-1293, USA
Address correspondence to A. S. Dolby, Department of Biology, University of South Florida, Tampa, FL 33620, USA. E-mail: dolby{at}chumal.cas.usf.edu .
Received 21 July 1998; revised 21 May 1999; accepted 23 June 1999.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Mixed-species flocks of birds form during winter in the eastern deciduous forests of North America. These flocks consist of two flock-leading nuclear species, tufted titmouse (Baeolophus bicolor) and Carolina chickadee (Poecile carolinensis), and several follower, or satellite, species, including downy woodpecker (Picoides pubescens) and white-breasted nuthatch (Sitta carolinensis). Hypotheses explaining the adaptiveness of participation in such mixed-species foraging groups have focused on increased foraging success and/or decreased predation risk. We tested the prediction that if nuthatches join nuclear species to reduce predation risk, they should be more reluctant to visit an exposed feeder in the absence of titmice than in their presence. When the feeder was positioned 16 m from forest cover, latency to visit the feeder was greater for both male and female nuthatches when titmice were absent. Removal of titmice had no effect on latency at 8 m. In the absence of titmice, nuthatches visited the feeder less frequently at both distances. These results indicate that reduced predation risk is a benefit that satellite species gain by flocking with nuclear species.
Key words: Baeolophus bicolor, group foraging, mixed-species group, predation, Sitta carolinensis, social behavior, tufted titmouse, white-breasted nuthatch.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Reduced predation risk and increased foraging success are thought to be among the selective advantages of social foraging (reviewed by Barnard and Thompson, 1985
;
Morse, 1977;
Moynihan, 1962;
Powell, 1985). Some of these
benefits extend to members of both monospecific and heterospecific groups, but
as monospecific group size increases, competition may begin to erode the
advantages of sociality (Alatalo,
1981; Barnard and Thompson,
1985). Members of mixed-species associations may be able to retain
the advantages of group foraging while being less subject to competition
(Barnard and Thompson, 1985;
Gartlan and Struhsaker, 1972).
Additionally, members of mixed-species groups may perform specialized roles,
such as providing alarm information about specific predatory threats
(Gautier-Hion et al., 1983;
Sullivan, 1984).
Heterospecific foraging groups occur during the winter within the community
of bark-foraging birds of eastern North America. Developing during early
autumn, these flocks are composed of two parid (family Paridae) nuclear
species, tufted titmouse (Baeolophus bicolor) and either Carolina
(Poecile carolinensis) or black-capped (P. atricapillus)
chickadee, and several satellite species including downy woodpecker
(Picoides pubescens) and white-breasted nuthatch (Sitta
carolinensis) (Morse,
1970). The parids seem to facilitate flock formation and to
initiate movements; the satellite species appear to follow the parids
(Morse, 1970).
The objective of the present experiment was to test the hypothesis that reduction of predation risk is a benefit that satellite species gain from foraging with nuclear species in these flocks. The primary predators of all flock participants are accipitrine hawks (Accipiter cooperii or A. striatus). We focus here on one satellite species, the white-breasted nuthatch, and one nuclear species, the tufted titmouse.
We predicted that when titmice were removed from woodlots, nuthatches would be more reluctant to visit a feeder positioned away from forest cover, where predation risk is presumably higher. More specifically, we predicted that nuthatches would exhibit greater delay before initially visiting the feeder and would visit the feeder less frequently in the absence of titmice than in their presence. Additionally, we predicted that the effect of titmouse removal on feeder visitation by nuthatches would become more pronounced if we positioned the feeder at greater distance from cover.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Study site and design
We conducted this experiment during the winter of 1997-1998 in 10 woodlots averaging 5.4 ± 0.5 ha (SD) in size in Union County, Ohio, USA. These deciduous forest fragments were completely surrounded by cultivated fields, and each contained one isolated mixed-species flock. Such flocks contained four to six chickadees, three to six titmice, four to six downy woodpeckers, and one nuthatch pair. Flocks were never observed crossing into neighboring woodlots.
We observed nuthatch feeder visitation under four treatment conditions using a balanced, crossover design with randomization. Nuthatches were randomly observed first either with titmice absent or with two titmice present. With titmice either present or absent, we randomly presented the feeder at 8 m or 16 m away from forest cover. Because we completed only 10 replicates, we were two woodlots short of a completely balanced design. However, our design was not compromised because the replicates we did carry out were randomly selected (Leighty R, personal communication).
Experimental protocol
Two weeks before the start of a given replicate, we placed a
radio-controlled sunflower seed feeder/trap
(Pierce and Grubb, 1981) and a
suet cage within a woodlot 10 m from its leeward edge. On the first day of the
replicate, we erected a metal fence post at the woodlot edge, as well as at 8
m and 16 m out into the adjacent field. Over the edge fence post, we placed a
6 cm x 2 m section of plastic pipe fitted with hooks to hold a feeder
and covered with bark to provide a landing substrate for birds. With this
arrangement, we could move the feeder by slipping the pipe off one fence post
and onto another. At about 20 m along the edge of the woodlot from the line of
fence posts, we positioned a burlap-covered blind measuring 1.0 x 1.5
x 1.5 m, from which we recorded behavioral responses. Finally, we moved
the feeder/trap to the fence post at the woodlot edge.
On the second day of each replicate, we removed the feeder/trap from the fence post and replaced it with a second feeder to be used in the experiment. This feeder resembled the feeder/trap, except that the trap component had been removed so that more than one bird could access the seed supply at any one time. We then conducted a training period during which we successively moved the feeder out to the 8 m and 16 m distances. We did not move the feeder until it had been visited by both nuthatches and by at least two titmice and two woodpeckers. After all target birds had visited the feeder at all three locations, we removed the feeder and placed the feeder/trap back at its original position 10 m inside the woodlot.
The following morning we captured birds using the trap and mist nets, leaving either two titmice and two to zero downy woodpeckers or zero titmice and two to four woodpeckers. Either two titmice or two woodpeckers were held in captivity so that they could be used in the subsequent trial. All chickadees and any additional titmice were removed and released 50 km away.
Three chickadees, one in each of three woodlots, evaded capture and remained in the woodlots throughout the experiment. The feeder was also occasionally used by American goldfinches (Carduelis tristis) and northern cardinals (Cardinalis cardinalis), species seldom occurring in mixed-species flocks with nuthatches.
The next morning, upon appearance of the nuthatches, we moved the feeder from the woodlot edge to either the 8 m or 16 m distance, waited 10 min, and then observed nuthatches for 45 min. We recorded for each gender the time delay until the initial trip was made to the feeder (latency), the number of trips made, and the total time spent away from forest cover. After 45 min, we moved the feeder to the other distance and conducted an identical observation session. All observations took place between 0800 and 1030 h.
At the end of the second 45-min observation session, we returned to the woodlot either the two woodpeckers or the two titmice that had been held in captivity. In exchange, two individuals of the opposite species were removed to control for the potentially confounding effects of group size. We then conducted a second day of observation using the same order of distance presentation as on the first day.
Analysis
Latencies to visit the feeder were examined using survival analysis
(Kleinbaum, 1996;
Parmar and Machin, 1995). For
each treatment-and-distance combination, we first constructed Kaplan-Meier
survival curves for time delays until the first trips to the feeder were made
(Statsoft, 1995). Curves were
constructed separately for each gender. We then used the logrank procedure to
test for differences between treatment groups (presence versus absence of
titmice) at each distance (Mehta and
Patel, 1995). We stratified our analyses by individual
(France and Lewis, 1991).
Numbers of trips nuthatches made to the feeder were analyzed using Poisson
regression (SAS Institute,
1992). The overdispersion parameter was used to account for
variability among subjects in our crossover design. We included feeder
distance, the presence or absence of titmice, and temperature in our model. We
added temperature as a covariate to control for its possibly confounding
effects on feeder visitation (Pravosudov
and Grubb, 1997). Sexes were analyzed separately. We used Spearman
correlation coefficients to determine whether numbers of trips made by
nuthatches to the feeder at each distance were correlated with numbers of
trips made by titmice and time spent at the feeder by woodpeckers. Results are
presented as mean ± SD, where applicable.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
The average time taken by males to complete a round-trip visit to the feeder was 7.8 ± 0.9 s (n = 10) at 8 m and 11.6 ± 1.0 s (n = 10) at 16 m. Analogous times taken by females were 7.2 ± 0.9 s (n = 10) at 8 m and 11.4 ± 0.5 s (n = 10) at 16 m.
Figure 1 displays the cumulative proportion of individuals delaying their first trip to the feeder placed at 8 m or 16 m from forest cover. When the feeder was positioned 8 m from the woodlot edge, neither male nor female nuthatches showed statistically greater hesitation to fly to it in the absence of titmice than in the presence of titmice (Table 1 and Figure 1). When the feeder was positioned at 16 m, male nuthatches delayed significantly longer when titmice were absent (Table 1 and Figure 1). Female nuthatches also tended to delay longer before visiting the feeder at 16 m when titmice were absent (Table 1 and Figure 1). Although this latter difference was not statistically significant, only 2 of 10 females visited the feeder at 16 m in the absence of titmice (Figure 1). This proportion was reversed in the presence of titmice. At both distances, more individuals of both sexes visited the feeder when titmice were present (Figure 1).
|
|
As noted above, 3 of the 10 woodlots contained a single chickadee that we
were not able to trap. Any bias due to the presence of these nuclear-species
individuals should have reduced chances of finding an effect of parid removal.
Nonetheless, to evaluate the impact of their presence during the experiment,
we reanalyzed the data using only the seven woodlots from which all of the
chickadees had been removed. Reanalysis revealed that when the feeder was
placed at 8 m, neither male (log-rank 
2 = 1.52, p
=.31) nor female (log-rank 2 = 1.83, p =.28)
nuthatches hesitated significantly longer before making their first trip when
titmice were absent, a result consistent with the analysis using all 10
woodlots. At 16 m, however, the effect of titmouse removal on latency to visit
the feeder was marginal for male nuthatches (log-rank 2 =
3.06, p =.06) and significant for females (log-rank 2
= 2.80, p =.03).
Males made more trips to the feeder during both 45-min observation periods when titmice were present (Figure 2), and this difference approached statistical significance (p =.06; Table 2). The number of trips made by males was not affected by either distance or temperature (Table 2), and neither the treatment x distance interaction nor the treatment x temperature interaction was significant (Table 2).
|
|
Females made significantly more trips to the feeder when titmice were present, but their frequency of visitation did not differ significantly between the two distances (Figure 2, Table 2).
We reanalyzed the numbers of trips made to the feeder by male and female nuthatches using only the seven woodlots that did not contain chickadees. This analysis produced no changes in statistical significance from the original analysis using all 10 woodlots.
The number of trips to the feeder by either male or female nuthatches at 8 m or 16 m was not significantly correlated with the total number of titmouse trips, nor with the amount of time spent by downy woodpeckers at the feeder (Tables 3 and 4).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
In a test where flock size was held constant, the tendency for white-breasted nuthatches to visit a feeder placed away from forest cover was affected by the presence of tufted titmice, as measured by both latency to visit the feeder and frequency of visits. These results support the hypothesis that reduction of predation risk is a benefit satellite species gain by flocking with nuclear species.
Consistent with our predictions, removal of titmice had a significant effect on delay to visit at 16 m, but not at 8 m. Although the total time spent away from cover differed by only 4 s between the two distances, the nuthatches seemed to be sensitive to such a small increase in exposure time. By contrast, removal of titmice had no differential effect on visitation frequencies of nuthatches at the two distances (i.e., no treatment x distance interaction). It is therefore difficult to assess the overall sensitivity of nuthatches to this difference in exposure time based on both response variables we measured.
At neither distance was the number of trips to the feeder made by
nuthatches correlated with the number of trips made by titmice. Gaddis
(1980) observed that in
mixed-species flocks in Florida, titmice or chickadees were the initiators of
alarm calls in response to the appearance of Cooper's or sharp-shinned hawks.
Furthermore, Sullivan (1984)
found that downy woodpeckers seemed to exploit information furnished by parid
alarm and contact calls. Transfer of alarm information among species has also
been observed in a number of other mixed-species groups
(Barnard and Thompson, 1985;
Gautier-Hion et al., 1983;
Hodge and Uetz, 1992;
Mathis et al., 1996;
Rasa, 1983). This behavior may
also be exhibited by white-breasted nuthatches. The observation that the
nuthatches seemed sensitive to the presence of titmice, but not to the
frequency of visitation of titmice to the feeder, suggests that nuthatches may
benefit from the presence of titmice even if the two species are in only
general proximity.
This experiment does not rule out other potential benefits to satellite
species from foraging with nuclear species. Nuthatches could gain foraging
advantages by copying (Krebs,
1973), kleptoparasitism
(Brockmann and Barnard, 1979),
or vigilance reduction (Elgar,
1989; Pulliam,
1973). Dolby and Grubb
(1998) found that both the male
and female of lone pairs of white-breasted nuthatches inhabiting isolated
woodlots from which parids had been experimentally removed exhibited elevated
vigilance rates compared to isolated pairs of nuthatches in control woodlots
with parids. Furthermore, in woodlots without parids, male nuthatches were in
reduced nutritional condition as indexed by ptilochronology
(Dolby and Grubb, 1998; Grubb,
1989,
1995). Mechanisms such as the
dilution and confusion effects may also reduce nuthatch predation risk in
mixed-species flocks (Hamilton,
1971). In conclusion, our experiment shows that the presence of
flock-leading species affects risk taking by nuthatches and suggests that
exploitation of titmouse vocal communication should be investigated further as
a mechanism favoring mixed-species flock participation by nuthatches.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
We thank the Crabbe, Geyer, Goddard, Mayers, Michaels, Petinski, Schmitter, Scott, Styre, and Thomas families for allowing access to their woodlots. P. F. Doherty Jr., J. D. Groom, and E. V. Pravosudova provided useful discussions. We also thank R. Leighty and T. A. Waite for statistical advice. The comments of G. G. Berntson, W. M. Masters, and T. A. Waite improved the manuscript. Financial support was provided by the Wilson Ornithological Society, The Ohio State University Alumni Association, and National Science Foundation grant IBN-9522064. This project was conducted under U.S. collecting permit PRT-673317, U.S. banding permit 20653, Ohio collecting and banding permit 650, and Ohio State University animal use protocol 94A0149.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Alatalo RV, 1981. Interspecific competition in tits Parus spp. and the goldcrest Regulus regulus: foraging shifts in multispecific flocks. Oikos 37:335-344.
Barnard CJ, Thompson DBA, 1985. Gulls and plovers: the ecology and behaviour of mixed-species feeding groups. London: Croom Helm.
Brockmann JH, Barnard CJ, 1979. Kleptoparasitism in birds. Anim Behav 27:487-514.
Dolby AS, Grubb TC, Jr., 1998. Benefits to satellite members in mixed species foraging groups: an experimental analysis.Anim Behav 56:501-509.[Web of Science][Medline]
Elgar MA, 1989. Predator vigilance and group size in mammals and birds: a critical review of the empirical evidence. Biol Rev 64:13-33.
France LA, Lewis JA, 1991. The analysis of failure time data in cross-over studies. Stat Med 10:1099-1113.[Medline]
Gaddis P, 1980. Mixed flocks, accipiters, and antipredator behavior. Condor 82:348-349.
Gartlan JS, Struhsaker TT, 1972. Polyspecific associations and niche separation of rain-forest anthropoids in Cameroon, West Africa. J Zool 168:221-266.
Gautier-Hion A, Quris R, Gautier J, 1983. Monospecific vs. polyspecific life: a comparative study of foraging and antipredatory tactics in a community of Cercopithecus monkeys. Behav Ecol Sociobiol 12:325-335.
Grubb TC Jr, 1989. Ptilochronology: feather growth bars as indicators of nutritional status. Auk 106:314-320.
Grubb TC Jr, 1995. Ptilochronology: a review and prospectus. Curr Ornithol 12:89-114.
Hamilton WD, 1971. Geometry for the selfish herd.J Theor Biol 31:295-311.[Web of Science][Medline]
Hodge MA, Uetz GW, 1992. Antipredator benefits of single- and mixed-species grouping by Nephila clavipes (L.) (Araneae, Tetragnathidae). J Arachnol 20:212-216.
Kleinbaum DG, 1996. Survival analysis, a self-learning text. New York: Springer-Verlag.
Krebs JR, 1973. Social learning and the significance of mixed-species flocks of chickadees (Parus spp.). Can J Zool 51:1275-1288.
Mathis A, Chivers DP, Smith RJF, 1996. Cultural transmission of predator recognition in fishes: intraspecific and interspecific learning. Anim Behav 51:185-201.
Mehta C, Patel N, 1995. StatXact-3 for Windows. Cambridge: Cytel Software Corporation.
Morse DH, 1970. Ecological aspects of some mixed-species foraging flocks of birds. Ecol Monogr 40:119-168.
Morse DH, 1977. Feeding behavior and predator avoidance in heterospecific groups. Bioscience 27:332-339.
Moynihan M, 1962. The organization and probable evolution of some mixed-species flocks of neotropic birds. Smithson Misc Coll 143:1-140.
Parmar MKB, Machin D, 1995. Survival analysis, a practical approach. Chichester, England: Wiley and Sons.
Pierce V, Grubb TC, Jr., 1981. Laboratory studies of foraging in four bird species of deciduous woodland. Auk 98:307-320.
Powell GVN, 1985. Sociobiology and adaptive significance of interspecific foraging flocks in the Neotropics.Ornithol Monogr 36:713-732.
Pravosudov VV, Grubb TC, Jr., 1997. Energy management in passerine birds during the nonbreeding season: a review. Curr Ornithol 14:189-228.
Pulliam HR, 1973. On the advantages of flocking.J Theor Biol 38:419-422.[Web of Science][Medline]
Rasa OAE, 1983. Dwarf mongoose and hornbill mutualism in the Taru Desert, Kenya. Behav Ecol Sociobiol 12:181-190.[Web of Science]
SAS Institute, 1992. SAS release 6.08. Cary, North Carolina: SAS Institute Inc.
Statsoft, 1995. STATISTICA 5.0 for Windows. Tulsa, Oklahoma: Statsoft.
Sullivan KA, 1984. Information exploitation by downy woodpeckers in mixed-species flocks. Behaviour 91:294-311.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M.M Webster, A.J.W Ward, and P.J.B Hart Shoal and prey patch choice by co-occurring fishes and prawns: inter-taxa use of socially transmitted cues Proc R Soc B, January 22, 2008; 275(1631): 203 - 208. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. V. Pravosudova and T. C. Grubb Jr. An experimental test of the prolonged brood care model in the tufted titmouse (Baeolophus bicolor) Behav. Ecol., May 1, 2000; 11(3): 309 - 314. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||