Search Tactics of Insectivorous Birds Foraging in an Australian Eucalypt Forest
Total Page:16
File Type:pdf, Size:1020Kb
SEARCH TACTICS OF INSECTIVOROUS BIRDS FORAGING IN AN AUSTRALIAN EUCALYPT FOREST RICHARD T. I-IOLMES 1 AND HARRY F. RECHER2 •Departmentof BiologicalSciences, Dartmouth College, Hanover, New Hampshire03755 USA, and 2Departmentof Vertebrate Ecology, Australian Museum, 6-8 CollegeStreet, Sydney South 2000, Australia A•STR^CT.--Thedifferent ways birds searchedfor food in an AustralianEucalyptus forest led them to detectand capturedifferent kinds of prey. Five major searchingmodes were identifiedamong 23 common,mostly insectivorous bird species.These were distinguished largelyby the rates,distances, and anglesmoved by birdswhile foragingand by their prey- capturebehavior. Some bird speciestypically moved slowly, visually examiningsubstrates at relatively long distances,and then took flight to captureprey (e.g. whistlers,flycatchers, muscicapidrobins, cuckoos). Others moved at more rapid rates and either gleanedsmall prey from nearby substrates(e.g. thornbills, treecreepers)or flushedinsects that were then pursued(e.g. fantails).Two species(Eastern Shrike-Tit, Falcunculus frontatus; White-eared Honeyeater,Meliphaga leucotis) were specializedsubstrate-restricted searchers, seeking in- vertebrateand carbohydratefoods among the exfoliatingbark of Eucalyptus. The searchtactics of birds in this south temperateAustralian forest were similar to those of birds in a north temperateforest in New Hampshire,USA, previouslyreported by Rob- inson and Holmes (1982). The differencesin food-searchingbehavior between these phy- logeneticallydistinct avifaunas (e.g. searchflight and prey-attackflight lengths,hop/flight ratios,foraging rates) reflect the effectsof unique foliagestructures (e.g. spacing of branches, arrangementsof leaves)and food resourcesat eachsite. Thesefindings support the hypothesisthat habitatstructure and food availabilityprovide opportunitiesand constraintson how birds searchfor and capturefood in foresthabitats. Thesein turn are postulatedto affectthe successof particularbird speciesexploiting those habitatsand thus influence bird community patterns.Received 25 October1985, accepted 25 January1986. TH• subjectof how and where birds obtain (1975) reached a similar conclusion from anal- their food has been central to the field of avian yses of forest structure and bird speciesdiver- ecology (Lack 1954, MacArthur 1958, Watson sity in tropical forests. 1970).While many studiesof bird foraging have Although food-searchingbehavior has been focused on prey-capture techniques and sub- analyzed frequently for species that feed in stratesfrom which prey are taken (e.g. Morse open country,such as ground-foraging thrush- 1968, Holmes et al. 1979, and others), recent es (Smith 1974a, b; Paszkowski 1982; Moreno findingshave shown that the methodsused by 1984) and aerial-feeding flycatchers (Davies birds in searchingfor food, leading up to prey 1977),only Morton (1980)and Fitzpatrick(1981) capture, and the factors that influence these in the Neotropics and Robinson and Holmes searchingpatterns may be particularly impor- (1982,1984) in the north TemperateZone have tant for understandingbird dietsand ultimate- analyzed quantitatively searching patterns of ly communitystructure. Robinson and Holmes forest birds. (1982, 1984) showed that search tactics of for- We studied the ways in which insectivorous aging birds in a north temperate forest were birds search for and capture prey in a south relatedto the physicaland biotic featuresof the temperateeucalypt forest in southeasternAus- environment, primarily vegetation structure tralia. We then comparedthe resultswith those and prey availability. These in turn were pos- of Robinson and Holmes (1982), who used the tulated to limit the foraging opportunities same methods and analytical procedures.The availableto birds in a given place and, hence, latter study was conducted in a temperate de- which speciesof birds could successfullyex- ciduous forest at Hubbard Brook in New ploit and survive in that environment. Pearson Hampshire, USA, where both the flora and avi- 515 The Auk 103: 515-530. July 1986 516 HOLMESAND RECHER [Auk, Vol. 103 fauna are phylogenetically distinct from those vation time was 20,262 s. For the more common in the Australian study area. Although there is species,20-30 different individuals were observed, no reason to expect direct niche equivalency while for the less common ones, at least 6 individuals between these geographicallyisolated and in- of each were representedin the sample. dependentlyevolved avifaunas,comparisons of Searchingmovements were divided into (1) flights, in which birds changeperches by flying, and (2) hops, bird foraging patterns in such contrastingen- in which they shift positionwithout extendingtheir vironments allow us to identify the habitat fea- wings. For all flights, we estimated the angles (in tures that influence bird search behaviors and increments of 45ø from the horizontal) and distances diets and thus to understand the factors that moved (in increments of 0.3 m for distances <1 m, help determine bird community structure. 0.5 m for those between 1 and 2 m, and 1-m units for longer flights). We recorded the frequency of hops but, because most were short moves of <5 cm, dis- STUDY AREA AND METHODS tancesmoved while hopping were not estimated. The study was conducted on the Southern Table- We recordedall attempts to capture prey and the lands of New South Wales (NSW) and adjacentVic- substrateson which the prey were located. Because toria in southeastern Australia between 1 October and we could not always determine if attemptswere suc- 31 December 1980. Three 10-ha study areaswere lo- cessful, any prey-directed action was considered a cated approximately 40 km southeastof Bombala, prey attack. When an attack involved a flight, we NSW (36ø54'S,149ø14'E) near the Bondi State Forest. estimated the angle flown and the distance moved. Detailed descriptionsof the vegetation,climate, and Prey attackswere recorded as glean, hawk, snatch bird populationsof the Bondi study areasare given (or hover), pounce,probe, or prise. Theseare defined by Recher et al. (1983, 1985) and Recher and Holmes and describedby Holmes et al. (1979) and Recher et (1985). For our purposes here, we treat these three al. (1985). plots as a single habitat, specificallya forest-wood- Foraging observationswere transcribedfrom the land ecotonegrading from a moist, tall, open forest tapes and timed using a stop watch. We acceptedfor through drier, open forest to woodland at the edge analysisonly sequences> 30 s in length, and for each of grazed pastures. of these we tallied the number of hops, flights, and The dominant trees in the moist forest were Euca- prey attacks.Dividing these by the length of each lyptus radiata,E. viminalis,and E. dalrympleana,with sequence,we obtained measuresof searchand prey- occasionalE. fastigataand E. cypellocarpa.Canopy attack rates. These rates were averaged over all se- height averagedabout 22 m, with a few trees reach- quences for each bird species, following Robinson ing 40 m. The subcanopy,shrub, and ground vege- and Holmes (1982). Correlations were based on val- tation was well developed and relatively dense. ues for individual species,not guild averages.Search tactics of Bondi and Hubbard Brook birds were com- The areas of drier forest were dominated by E. ra- diata,E. viminalis,E. ovata,and E. dalrympleana.Can- pared, where possible,using Mann-Whitney U-tests. opy height averaged 16-20 m. The shrub layer was RESULTS AND DISCUSSION relatively open, due to the grazing of domesticstock and somelogging. The dry forest graded into wood- THE BIRD COMMUNITY AND FORAGING GUILDS land dominatedby E. paucifioraand E. stellulata,which had a lower canopy(8-10 m) than the forest. We obtained data on the food-searchingpat- We quantified the behavior of foraging birds at Bondi using the methods of Robinson and Holmes terns of 23 speciesof insectivorous birds [see Table 1 for scientific names and taxonomic (1982). We moved systematicallyabout each study plot, and when a foragingbird was encountered,re- (family) affiliations].These representthe most cordedits actionson a tape recorder.Only adult birds common, relatively small birds (7-76 g) that that were clearly foraging were followed, and these feed primarily on insectsand other arthropods were observedfor as long as they could be kept in in eucalyptforests of southeasternAustralia (see sight. When the bird was lost from view or when it Loyn 1980;Recher et al. 1980,1985; Recher and stoppedforaging, we moved on until another active Holmes 1985 for details on the forest avifauna forager was located. We alternated study plots and of this region).We includedas insectivores the routes between days and at different times of day. Yellow-faced, White-eared, and White-naped The relatively open and evenly dispersedfoliage of these eucalypt habitats permitted good visibility at honeyeaters(Meliphagidae), which are often all levels of the vegetation. considered nectar feeders (Recher et al. 1985) We obtained data from 1,136 foraging sequences but which take most of their food from the sur- on 23 bird species;the averagelength of foraging faceof foliageor from under loosebark (Table sequenceswas 79 s, and the total accumulatedobser- 1). Observationson the searchbehavior of oth- July1986] SearchTactics ofAustralian Birds 517 er less common speciesalso were gathered and Frequencydistributions and averagelengths are referred to below where appropriate. of search flights differed among speciesand The 23 specieswere divided into