174 SHORT COMMUNICATIONS mained nesting. However, productivity in 1995 was We thank the Minerals Management Service for higher than in 1996 (D. G. Roseneau, pers. comm.), their support,M. F. Chance and l? E Chance from the and 1996 may have been a more stressfulyear. Alaska Maritime Refuge for their valuable advice and Murres are probably the deepest diving alcid (Piatt logistical help, and _I. W. Hupp, T. S. Smith, and W. and Nettleship 1985), and alcids in general have higher D. Ostrand for their manuscriptreviews. wing loads than other flying seabirds (Pennycuick 1987). Either of these factbrs

The Condor 100~174-177 0 The Cooper Ornithological Society 1998

CORRELATES OF CREEPING SPEED VARIABILITY IN TWO SPECIES OF l

TOMASZ S. OSIEJLJK Departmentof Morphology, Universityof A. Mickiewicz, Szamarzewskiego91A, 60-569 Poznari, Poland, e-mail: [email protected]

Abstract. Foraging behavior of Short-toed Tree- supportedthe model predictions.Short-toed creeper bruchydactylaand EurasianTreecreeper crept, on average, slower than . C. familiaris was examined to test how well it fits a Both species crept more slowly on larger trees. Eur- simple foraging model using data on arthropoddistri- asian Treecreeper crept slower on thick English oaks bution on tree trunks. Field observations in general than on thick Scotch pines. Both treecreepersmoved slower when probing as compared to , which was expected as foraging technique strongly affects ’ ’ Received 8 April 1996. Accepted 17 October 1997. handling time. SHORT COMMUNICATIONS 175

Key words: Certhia brachydactyla, Certhia fami- aging on oaks and pines where initial and final height liaris, foraging behavior, optimal foraging hypotheses, of foraging were different. There were two reasons for treecreepers. this: (1) observations of birds foraging with no move- ment (H, = HZ) were often very short and may have been influenced by the observer and (2) sufficiently Treecreepers, Certhia spp., are useful subjects for detailed data concerning abundance were modeling and testing optimal foraging hypotheses be- available only for pine and oak. Observations were cause of their simple foraging behavior. Treecreepers’ obtained for about 30 individual Eurasian Treecreepers foraging behavior on tree trunks consists of many rep- and about 50 individual Short-toed Treecreepers each etitions of search-encounter-decide (Stephens and year. Abundance was estimated on the basis of regular Krebs 1986). Treecreepers take prey of rather uniform mapping in 1991 and 1992 (Osiejuk 1993) and mark- size, especially in comparison to their body size (Kui- inz birds on the same studv olot in 1996 and 1997 tunen and Tormall 1983, Kuitunen 1989, Suhonen and afkr this study was completed iosiejuk and Kuczydski Kuitunen 1991). Therefore, we can assume that han- 1997, and unpubl. data). Consequently, pseudorepli- dling time for all prey is similar, depends mainly upon cation of observations was minimal. the foraging technique used, and is longer for probing I distinguished two patterns of movements along the than for gleaning (Osiejuk 1994, 1996). trunk in treecreepers: straight and spiral (Osiejuk If, following Holling (1959), T, is the time spent 1996). When the spiral pattern is observed, the real searching and T, is the time spent handling, then T, + distance of movement is greater than obtained on the T, = T, where T is all the time spent foraging on one basis of H, - H,. The percentage of spiral pattern use tree. If we also assume that the speed of movements was similar in C. brachvdactvla and C. familiaris (34% between encounters is constant, the mean speed of and 39%, respectively,‘X2, =’ 2.7, P >“0.09). _ creeping should decrease with increasing prey density. Data on arthropod assemblage on the surface of tree Therefore, based upon these assumptions and data con- trunks were taken from two earlier studies. Jackson cerning food abundance on tree trunks, predictions (1979) showed that prey density and availability de- about creeping speed on different tree species or trees pend upon the structure of bark, which varies accord- with different diameter should be possible. However, ing to tree species and age. Therefore, one could ex- most observational methods that have been used do pect that trees with greater diameter-that is, older, not allow estimation of creeping speed on different tree with more complicated bark surface-should be richer species (Vanicsek 1985, SzCkely 1987, Suhonen and in arthropod supply than thinner trees of the same spe- Kuitunen 1991). This paper presents predictions of a cies. Dziabaszewski (1976) defined auantitativelv and simple theoretical foraging model and compares them qualitatively how abundance of Araihnoidea changes with creeping speed patterns in two treecreeper spe- in the course of the year on different tree species and cies. on different parts of the tree. He also compared the abundance of Arachnoidea and Insecta giving com- METHODS plete and reliable data for testing the above mentioned Field work was carried out during five winters between model. Dziabaszewski (1976) confirmed that there is 1990 and 1995 in Wielkopolski National Park a significant difference in arthropod distribution on (52”15’N, 16”50’E), western Poland. The foraging be- Scotch pines and English oaks. The overwinter abun- haviors of Short-toed Treecreeper Certhia brachydac- dance of on oaks depends mainly upon the tyla and Eurasian Treecreeper C. familiaris were ob- initial number of and on the whole tree. served between 08:OO and 12:00, from 1 December to Each winter the majority of arthropods is forced to 28 February. To minimize the effect of weather, ob- migrate from the tree canopy down to the trunk with servations were made only on good days (temperature numerous crevices in its bark, because dry leaves do over -SC, wind below 4” in Beaufort scale, no rain not protect them from frost. In the case of pines, the or snow). Observations, which were taken during extent of migration is much more dependent upon the walks in the study area, were not started where the weather because twigs with needles are a sufficient birds were sighted but after their first flight among the refuge during mild winters. Generally, arthropods were trees. A maximum of three observations were made always more abundant on oaks than on pines. Dzia- per individual before another bird was located. baszewski (1976) observed on average of 57 arthro- Amount of double and triple observations of the same pods m-2 of Scotch pine bark and 186 mm* of English individual was less than 2% of the data set in both oak bark. species. The following foraging variables were recorded and MODEL AND PREDICTIONS constituted any single observation: (1) tree species The model I consider is intended to be the simplest used for foraging (Scotch pine Pinus sylvestris, En- one possible that incorporates parameters of foraging glish oak Quercus robur, other), (2) tree diameter (cm) place and foraging behavior and allows prediction of at breast height, (3) foraging technique (gleaning or creeping speed pattern of treecreepers. I made the fol- probing), (4) height (m) of beginning and finishing of lowing predictions: foraging on the tree (H, and H,, respectively), (5) movement patterns (straight or spiral), and (6) foraging (1) Within the same tree species, creeping speed time. More about observational methods and the study should be slower on trees with larger diameter as bark area can be found in Osiejuk (1994, 1996). Creeping of thicker trees are more cracked and give more hiding speed (m mini) was estimated as S = (H, - H,)/T. places for arthropods (Jackson 1979). I analyzed only the observations of treecreepers for- (2) The creeping speed should be lower on oaks than 176 SHORT COMMUNICATIONS

TABLE 1. Mean (2 SD) creeping speed (m min’) 2.34 m min’, using probing 2.22 2 1.25 m min-I; tjX4 in treecreepers with respect to tree species, tree di- = 3.7, P < 0.001). However, Short-toed Treecreeper ameter, and movement pattern. used probing technique rarely; the difference in pro- portion of probing and gleaning used was significant (x2, = 3.9, P < 0.05). This also is supportedby the results of an ANCOVA with tree species, foraging technique, and movement pattern as factors and tree Pine 540 Straight 4.2 ‘- 2.9 4.0 ? 2.3 diameter as covariate. There was a significant move- >40 3.3 2 0.2 4.0 ? 2.1 ment pattern effect on creeping speedwhen controlling Oak 540 Straight 4.3 2 1.5 3.7 ? 2.6 for the effect of tree diameter in Short-toedTreecreeper >40 2.8 2 1.6 4.1 ? 2.0 (ANCOVA; tree species, F,,,,, = 0.8, P > 0.4; for- Pine 540 Spiral 2.3 f 0.5 3.7 ? 0.2 aging technique, F,,,,, = 1.5, P > 0.1; movement pat- >40 1.5 ” 0.1 3.8 ? 0.2 tern, F1,Z86= 32.4, P < 0.001; covariate tree diameter, Oak 540 Spiral 2.6 + 0.6 2.7 5 0.9 F 1.286= 15.8, P < 0.001). In the case of EurasianTree- >40 1.4 + 0.1 1.3 2 0.1 creeper, there were significant effects of all factors on creeping speed when controlling for the effect of tree diameter (ANCOVA; tree species,F,,237 = 29.3, P < 0.001; foraging technique, F,,,,, = 41.6, P < 0.001; on pines becauseoak bark provides a richer arthropod movement pattern, F,,,,, = 31.2, P < 0.001; covariate supply. tree diameter, F,,,,, = 5.5, P < 0.05). (3) Creeping speed should be lower if probing is used, because this technique lengthens handling time DISCUSSION in comparisonto gleaning (Holling 1959, Stephensand The pattern of treecreepers’ speed of movement in Krebs 1986). C. brachyductyla probed less than half general supports the prediction of Holling’s equation as often (about 4% of feeding time) as C. familiaris (1959). The crucial factor influencing creeping speed (x2, = 3.9, P < 0.05). Therefore, differences in speed in this model is handling time, which is determinedby caused by gleaning:probinguse ratio should be more foraging technique. In both species,the foraging tech- apparentin C. familiaris. nique used strongly affected creeping speed. Short- RESULTS toed Treecreeper,however, used probing significantly less frequently than Eurasian Treecreeper.This differ- A total of 560 observationswas collected for C. bruch- ence and the significantly longer bills of Short-toed ydactyla and 420 for C. familiaris. Mean (? SD) du- Treecreeper (bill length in the study area, C. brachy- ration of observationswas 74 t 98 set for C. bruch- ductyla = 15.4 2 1.4 mm, C.fumiZiaris = 13.1 + 0.9 yductylu and 67 t 70 set for C. familiaris. Duration mm, t,, = 3.9, P = 0.001, Osiejuk and Kuczynski, of observationsdid not differ between species (t9,* = unpubl. data) may be very important evolutionarily. 1.4, P > 0.1). C. bruchyducryla crept slower than C. Among superficially dispersedprey that are equally familiaris (mean ? SD = 2.95 ? 2.24 m min’ and available for both treecreepers, there are many prey 3.61 ? 3.55 m min’, respectively; t978= -3.5, P < hidden in crevices of different depth. One can easily 0.001). If only observationsof foraging where H, # imagine that in some cases Eurasian Treecreepersare H, were considered,then creeping speed was 3.75 ? forced to stop for a moment and use probing to catch 1.84 m min-’ and 4.33 -C 3.46 m min-I, respectively; an arthropod, whereas Short-toed Treecreeper can &789= -3.0, P < 0.01). The correlation between di- catch an arthropod by gleaning. Therefore, the Short- ameter of tree trunk and creeping speed was negative toed Treecreeperprobably is better adapted to exploit and significant for both species (C. bruchyductyla, r, deeper bark crevices, and this adaptationcould be cru- = -0.24, n = 442, P < 0.001; C. familiaris, r, = cial in severe winters. Studies on the distribution of -0.32, n = 351, P < 0.001). Certhia species/subspeciesthroughout Eurasia in re- The comparisonsof mean creeping speed in Table gard to climate, forest type, and bill morphologywould 1 suggestthat there are factors, other than tree species, be valuable. tree diameter, and movement pattern, for which obser- vations differ from the model predictions. Short-toed The manuscriptwas greatly improved by comments Treecreepercrept distinctly faster on trees with smaller from Lechu Kuczynski, Piotr Tryjanowski, and two diameter, and in most casessuch differences were sig- anonymous referees. I thank Andrzej Dziabaszewski nificant (Table 1). In Eurasian Treecreeper the result for pricelessinformation about winter behavior. was similar in only one case (Table 1). The differences English usage was improved by Bruce Peterson between creeping speed on different tree specieswith through the Association of Field Ornithologists.I was similar diameter did not always confirm model predic- financially supportedby the Polish Scientific Commit- tions. There was no significant difference in creeping tee KBN, grant 6/P205/103/06. speed on pine and oak in Short-toed Treecreeper.Eur- asian Treecreeperfits the model better as in most cases LITERATURE CITED it crent slower on oak than on nine (Table 1). Both treecreeperscrept slower when used probing DZIABASZEWSKI, A. 1976. Aruchnoidea (Arunei, Opi- than gleaning, and in both cases the differences were liones, Pseudoscorpionida) on crowns of trees.An significant (C. bruchyductylu using gleaning 3.81 ? ecological faunistic study. Univ. A. Mickiewicz 1.85 m min-I, using probing 2.07 ? 0.52 m min’; tdZ5 Press, Poznan, Poland (in Polish with English = 4.0, P < 0.001; C. familiaris using gleaning 4.11 * summary). SHORT COMMUNICATIONS 177

HOLLING,C. S. 1959. Some characteristicsof simple behavior of the Den- types of predation and parasitism. Can. Entomol. drocopos major during winters with rich crops of 91:385-391. Scotch pine cones. Ornis Fennica 71:144-150. JACKSON,J. A. 1979. Tree surfaces as foraging sub- OSIEIUK,T S. 1996. Locomotion patternsin wintering stratesfor insectivorousbirds, p. 69-99. In J. G. bark-foraging birds. Omis Fennica 73:157-167. Dickson, R. N. Conner, R. R. Fleet, J. A. Jackson, OSIEJUK,T S., ANDL. KUCZY~~SKI.1997. Factors af- fecting song-ratein treecreeoersCerthia SVV. Adv. and J. C. Kroll [eds.], The role of insectivorous &. birds in forest ecosystems.Academic Press, New in Et&log; (Suppl.) 32: 129. York. STEPHENS.D. W.. AND J. R. KREBS. 1986. Foraaina KUITUNEN,M. 1989. Food supply and reproductionin theor’y. Princeton Univ. Press, Princeton, NJ. y y SUHONEN,J., AND M. KUITUNEN.1991. Intersexualfor- the common treecreeper(Certhiafamiliaris). Ann. aging niche differentiation within the breeding Zool. Fennici 26:25-33. pair in the Common TreecreeperCerthia familiur- KUITUNEN,M, AND T. TORMALA. 1983. The food of is. Omis Stand. 22:313-318. Treecreeper Certhia f familiaris nestlings in SZ&ELY, T 1987. Foraging behavior of woodpeckers southernFinland. Omis Fennica 60:42- 44. (Dendrocopos spp.), (Sittu europaea) OSIFJUK,T. S. 1993. Habitat selectionvs. territory size and treecreeper (Certhiu sp.) in winter and in in woodpeckers,treecreepers and nuthatch.M.Sc. spring. Ekol. Pol. 35101-114. thesis, Univ. A. Mickiewicz, Poznad, Poland. VANICSEK,L. 1985. The study of bird speciesforaging OSIWUK,T S. 1994. Sexual dimorphism in foraging on the bark. Aquila 95:83-93.

The Condor 100:177-179 0 The Cooper Omithologlcal Society 1998

INTERACTIONS BETWEEN BLACK-BILLED MAGPIE AND FALLOW DEER’

PETER V. GENOV Bulgarian Academy of Sciences, Institute of Zoology, Boul. T. Osvoboditel I, ZOO0 So$a, Bulgaria PAOLAGIGANTESCO Dipartimento di Scienzede1 Comportamento Animale e dellUomo,’ Universita ’ di Pisa, Via A. Volta 6, SSZOO, Pisa, Italy

GIOVANNA MASSEI Institute of Terrestrial Ecology, Hill of Brathens, Banchory, AB31 4BY, Aberdeen, UK, e-mail: [email protected]

Abstract. Black-billed Magpies (Pica pica) were wole 1965), or removing their ectoparasitesor fur (Is- observed pecking on fallow deer (Dama dama) on 56 enhart and DeSante 1985, Yosef and Yosef 1991, Fitz- occasions. Ectoparasite removal was apparently the patrick and Woolfenden 1996). The behavior of un- reason for this interaction. Birds preferred deer that gulates in response to birds landing on them ranges were sitting to deer that were standing,and interacted from adopting posturesthat facilitate parasiteremoval, preferentially with adult males over females or calves. to intolerant reactions. Black-tailed deer (Odocoileus Deer did not solicit cleaning and, on a few occasions, hemionus) freeze when cleaned by Scrub-Jays(Aphel- were observedto shakeoff birds. This interactionmay ocoma coerulescens) (Isenhart and DeSante 1985), be beneficial for magpies, becauseectoparasites are a whereas feral hogs and wild boar (both Sus scrofu) predictablesource of food, but its effect on fallow deer solicit cleaning by lying down when pecked by Com- remains to be investigated. mon Crows (Corvus brachyrhyncos) and Black-billed Key words: bird-ungulate interactions, Black- Magpies (Pica pica) (Kilham 1982, Massei and Genov billed Magpie, fallow deer, Pica pica. 1995). In contrast, oxpeckers sometimes induce horn- shakingby their hosts,possibly because oxpeckers pre- Several studies on large herbivore-bird interactions vent wound healing when removing ectoparasites have describedbirds using ungulatesas perches(Heat- (Watkins and Cassidy 1987). Black-billed Magpies eat ticks from the back of elk (Cervus canadensis) (Linsdale 1946), but have not pre- ’ Received 28 July 1997. Accepted 17 October viously being described to interact with fallow deer 1997. (Dama duma). A significant number of opportunistic