Intra- and Interspecific Calling in a T ropical Community

Paula L. Enríquez and J. Luis Rangel Salazar 1

Abstract.—We studied the intra- and interspecific r esponses to playback of pre-recorded calls by five tropical humid for est owl species at La Selva preserve in northeaster n fr om April to September 1995. Response to conspecific br oadcast calls differed among species (X2 = 24.4; df = 1; P < 0.001): V ermiculated Scr eech- owls (Otus guatemalae) responded to 47.6 per cent of br oadcasts, followed by Crested (Lophostrix cristata, 45 percent), Mottled Owls (Ciccaba virgata, 18.3 per cent), and Black-and-white Owls ( C. nigrolineata, 9 percent). Cr ested Owls (x = 16.37, sd = 2.6), Mottled Owls (x = 11.7, sd = 7.1), and Vermiculated Scr eech-owls (x = 10.9, sd = 0.9) responded to interspecific playback mor e than did Black- and-white Owls (x = 2.04, sd = 2) (H = 10.6; P = 0.01). Spectacled Owls (Pulsatrix perspicilata) did not r espond at all during our br oad- casting period. Both types of r esponse showed some monthly varia- tion. Response to the calling of other owls also depended on ecologi- cal variables such as selection, population density, and resource use. Our data suggest that the development of r elationships within the tr opical owl community at La Selva may have been medi- ated in part by intra- and interspecific calling.

The interactions between owl species in tr opical STUDY AREA communities have been poorly studied. In norther n Eur ope, the behavioral ecology of owl La Selva Biological Station is located in communities has r eceived some attention Sarapiquí county, Her edia province, Costa Rica (Korpimäki 1987), but no community-based owl (10˚26’N 83˚59’W). The station adjoins Braulio research has been conducted in the Neotr opics. Carrillo National Park to the south and agricul- Single-species r esearch has examined the food tural land and cattle pastur e to the north. La habits (Ger hardt et al. 1994a) and breeding Selva encompasses 1,513 ha, and main habi- biology (Gerhardt et al. 1994b) of Neotr opical tats are primary humid for est, young second owls, and the calling behavior of one wide- growth, grassland in the pr ocess of succession, spread species (Gerhardt 1991). Community abandoned plantations, swamps, for est study ecology is central to understand factors that plots, and open ar eas with buildings. Elevation regulate the structur e, dynamics, and evolution ranges from 35 to 150 m. W eather conditions of owl populations (Pianka 1988), and the are very humid, with 4,000 to 4,500 mm of effects of an inter - and intra-species interac- annual pr ecipitation. Annual temperatur es tion, and eventually, the conservation of those range from 24.7 to 27.1˚C. communities. In this paper we r eport the calling interactions among five Neotr opical METHODS rainforest owls from La Selva Biological Station, Costa Rica. Intra- and interspecific interactions among five owl species; Vermiculated Scr eech-owl (Otus guatemalae), Crested Owl (Lophostrix cristata), Spectacled Owl (Pulsatrix perspicilata), Mottled 1 Associate Researchers of Or nithology, El Owl (Ciccaba virgata), and Black-and-white Owl Colegio de la Frontera Sur , Apdo. 63, San (C. nigrolineata), were studied thr ough br oad- Cristóbal de Las Casas, Chiapas 29290, casting of vocalizations fr om April to September México. 1995. Br oadcasting has been a consistent

525 2nd Owl Symposium method for surveying woodland owls in North Kruskal-Wallis test (H) to compar e multiple America (Ganey 1990, McGarigal and Fraser intra- and interspecific r esponse percentages, 1985, Mosher et al. 1990), and the Neotr opics Kolmogor ov-Smir nov two sample test (D) for (Enríquez 1995, Ger hardt 1991). testing distributions of species r esponses to coexisting species and r esponses of coexisting At La Selva, and prior to our br oadcasting species to the tar get species, and Shapir o-Wilk period, we recorded the typical vocalizations of statistic (W) to test that interspecific r esponses the five owl species studied. An Uher 4000RL followed a normal distribution. and an Electr ovoice Unidir ectional Micr ophone were used to r ecord the vocalizations. Once we RESUL TS had obtained quality r ecordings, they were copied onto independent cassettes for each owl The owl community in La Selva and the sur - species. Closed loop cassettes used consisted rounding ar ea are represented by eight species; of 3 minutes of typical vocalizations; “hoot” five are listed in table 1 and, the Bar n Owl rates differed slightly among species. (Tyto alba), Least Pygmy-owl (Glaucidium minutissimum), and Striped Owl ( Asio clamator). We established 30 survey stations on thr ee During our br oadcasting period we taped 340 major trails (10 stations/trail). The trails broadcast vocalizations. An intraspecific selected covered proportionally all at response was obtained 82 times (25 per cent of La Selva (P > 0.05, Enríquez 1995). Starting total br oadcasts), whereas an interspecific points at each trail wer e at least 400 m apart. response was obtained 110 times (32.3 per - Survey stations along trails wer e 200 m apart cent). The Spectacled Owl did not r espond at from each other following Forsman (1983). The all during our br oadcasting period. sampling period was 10 minutes at each station, consisting of 3 minutes of br oadcasting Intraspecific Responses followed by 7 minutes of listening. Thr oughout this period, we noted the vocalizations of any Most of the owl species at La Selva r esponded owl that responded. To avoid provoking differ- generally more to the br oadcasting of conspe- ent species at the same station (Kochert 1986), cific vocalizations (table 1). Per centages of we randomly selected the br oadcast order of intraspecific r esponses differed among owl the five species and played the vocalizations of species ( X2 = 24.4, df 1, P < 0.001). The only one species at each station. Then, we Vermiculated Scr eech-owl and Cr ested Owls selected a new broadcast order twice on each had higher per centages of intraspecific r e- trail. We conducted surveys on each trail twice sponses than Mottled and Black-and-white monthly (N = 36). No surveys wer e conducted Owls (table 1). during rain. Sampling methods ar e described in mor e detail in Enríquez (1995). Collected Among those owl species that r esponded data were analyzed using a Chi-square test (X2) intraspecifically, variation on monthly per cent to evaluate the intraspecifc r esponse, a

Table 1.—Total mean of response percentages to broadcasting of pre-recorded vocalizations of five species of tropical humid forest owls at La Selva preserve in northeastern Costa Rica, from April to September, 1995.

Broadcasting Vocalization by1 Responses by1 VSO CO SO MO BWO

VSO 47.62 11.51 10.10 11.36 10.67 CO 16.89 45.11 16.63 18.55 13.42 SO MO 7.19 6.95 15.10 18.29 17.63 BWO 3.05 5.11 9.01

1 VSO (Vermiculated Screech-owl), CO (Crested Owl), SO (Spectacled Owl), MO (Mottled Owl), BWO (Black-and- white Owl). 526 of responses was distributed nor mally Crested Owls’ broadcast vocalizations (D = (Shapiro-Wilks test, P > 0.05). Cr ested and 1.41, P < 0.05). W e did not find this behavior Vermiculated Scr eech-owls showed more for any other owl species surveyed (P > 0.05). monthly variation in intraspecific r esponse Monthly variation of interspecific r esponses is than did Mottled and Black-and-white Owls (H shown in figur es 2-5. Vermiculated Scr eech- = 15.05, df 3, P < 0.05) (fig. 1). owls showed some monthly variation in r e- sponse to Black-and-white Owl vocalizations (W Interspecific Responses = 0.659, P < 0.05) (fig. 3). Also, Black-and- white Owls exhibited monthly variation in their Species that r esponded to calls of all other response to vocalizations of Cr ested Owls (W = sympatric species included in this survey wer e 0.678, P < 0.05) and Spectacled Owls (W = 0.7, Crested Owl, Vermiculated Scr eech-owl, and P < 0.05) (fig. 5). Mottled Owl. Meanwhile, the Black-and-white Owl responded to only two sympatric species DISCUSSION the Cr ested Owl and the Spectacled Owl (table 1). Cr ested Owls responded mor e frequently to Responses to br oadcast of conspecific calls interspecific br oadcast vocalizations (x = 16.37, were more frequent than interspecific r e- sd = 2.6) than did Mottled Owls (x = 11.7, sd = sponses at La Selva for two of five species 7.1), Vermiculated Scr eech-owls (x = 10.9, sd = studied. Our r esults suggest that Cr ested, 0.9), or Black-and-white Owls (x = 2.04, sd = Vermiculated Scr eech-, and Mottled Owls wer e 2) (H = 10.6; P < 0.01). On the other hand, we more responsive, both intra- and interspecific- did not find dif ferences in r esponse to any one ally, than Black-and-white and Spectacled of the five owl vocalizations br oadcast (H = Owls. In temperate for est, interspecific r e- 2.06; P = 0.72). sponses also varied among owl species, mostly during the br eeding season (Bosakowski et al. Crested Owls responded mor e frequently to 1987, Springer 1978). But, Smith et al. (1987) other species than other species r esponded to found that the Easter n Scr eech-owl (0tus asio)

Figure 1.—Monthly responses to intraspecific broadcasting of pre-recorded vocalizations of four species of tropical humid forest owls at La Selva preserve, northeastern Costa Rica, in 1995.

527 2nd Owl Symposium

Figure 2.—Crested Owl monthly responses to interspecific broadcasting vocalizations of co-existing species of tropical humid forest owls at La Selva preserve, northeastern Costa Rica, in 1995.

Figure 3.—Vermiculated Screech-owl monthly responses to interspecific broadcasting vocalizations of co-existing species of tropical humid forest owls at La Selva preserve, northeastern Costa Rica, in 1995. 528 Figure 4.—Mottled Owl monthly responses to interspecific broadcasting vocalizations of co-existing species of tropical humid forest owls at La Selva preserve, northeastern Costa Rica, in 1995.

Figure 5.—Black-and-white Owl monthly responses to interspecific broadcasting vocalizations of co- existing species of tropical humid forest owls at La Selva preserve, northeastern Costa Rica, in 1995. 529 2nd Owl Symposium responded less during the br eeding season. In La Selva responded to all other species in- a tropical for est in , Ger hardt (1991) cluded in this survey. Meanwhile, Black-and- reported that Mottled Owls r esponded to 40 white Owls responded to only two owl species. percent of br oadcasts during the br eeding Crested Owls responded more frequently to season. Mottled Owls r esponded less fre- interspecific br oadcast vocalizations than did quently in our study site than they did in Mottled Owls, Vermiculated Scr eech-owls, and Guatemala. Cr ested and Vermiculated Black-and-white Owls. The observed dif fer- Screech-owls responded to over 45 per cent of ences in interspecific r esponses at La Selva conspecific br oadcasts. The differences in the may be related to differences in population percentage of responses to br oadcasting could density, habitat use, and food habits. Also, be related in part to the species abundance in calling behavior of for est owls is affected by the study area. Enríquez (1995) found that environmental variables (Carpenter 1987), as Crested and Vermiculated Scr eech-owls were observed at La Selva (Enríquez 1995). quite abundant and Spectacled Owls less abundant at La Selva. Lack of r esponses by Crested Owls responded more to Mottled Owl Spectacled Owls and few responses by Black- vocalizations than other species pair combina- and-white Owls could be associated with their tions. These species used the same habitat at home range size. In Guatemala, a single Black- La Selva (Enríquez 1995). W e found these and-white Owl had a home range 20 times species calling together in dif ferent vegetation larger than Mottled Owls had (Ger hardt et al. strata. Mottled Owls feed on vertebrates like 1994b). small rodents, but ar e considered mainly insectivor ous (Ger hardt et al. 1994a). Possibly Habitat influences the abundance of some these species differ in feeding time, strategies species (Will 1986). For ests and old second and sites. On the other hand, the Mottled Owl growth habitats at La Selva favor the Cr ested is mor e tolerant of habitat change and so is Owl. This rar e species is probably a relict of a both abundant and br oadly distributed group that spr ead around the tr opics and (Mikkola 1992). Also, Mottled Owls can visit survived in old for ests with small changes urban ar eas to feed. through the period of climatic cooling (Hekstra 1973). The Vermiculated Scr eech-owl and Ciccaba owls showed also a high level of inter - Mottled Owl ar e widely distributed and com- action, mostly when Mottled Owls r esponded to mon in Neotr opical for ests (Gerhardt 1991, Black-and-white Owl calls. These Ciccaba Stiles and Skutch 1989). On the other hand, species overlapped in distribution and activity Spectacled Owls use open habitats with nearby patterns, and several times we listened to both woodlots to vocalize, hunt, r oost, and br eed. species calling simultaneously. Although these The Black-and-white Owl is rar e thr oughout its species both took lar ge numbers of insects, the entire distribution and may not use a particu- mammalian part of their diet showed little lar habitat at La Selva (Enríquez 1995). W e overlap (Gerhardt et al. 1994a), they likely used recorded this species in dif ferent habitats, different foraging strategies and captur e tech- calling and pr obably hunting. Nicholls and niques, and they used quite dif ferent br eeding Warner (1972) mentioned that although owl sites (Gerhardt et al. 1994b). species may use one habitat mor e than an- other, the habitat that is used less may not be In or der of size, Vermiculated, Mottled, and less important, since it could contain r esources Crested Owls feed mainly on invertebrates critical to the species’ survival. (Hekstra 1973), Black-and-white and Spec- tacled Owls feed mainly on vertebrates (Ibañez We did not find seasonality in owl r esponses in et al. 1992). Insects, caterpillars, crabs, mam- this study; Spectacled Owls, however, vocalized mals, bir ds, and reptiles were reported as prey from January to Mar ch only. Ther efore this for Spectacled Owl (Mikkola 1992). Johnsgar d species had a seasonality to its calling behavior (1988) mentioned that owl species that feed and our surveys were conducted outside the mainly on insects have small territories. For season during which it was most vocal. those owl species that have similar diets, competition can be r educed by utilizing dif fer- Interspecific r elations may include overlap in ent time or space. For Cr ested and Vermicu- distribution, hunting period, habitats, and food lated Screech-owls, the habitat most utilized (Mikkola 1983). Thr ee species (Crested Owl, was the cacao orchard, but these species Vermiculated Scr eech-owl, and Mottled Owl) at occupied dif ferent vegetation strata. W e found 530 Crested Owls calling in the canopy and V er- and conservation of norther n forest owls: miculated Scr eech-owls were in the understory. symposium pr oceedings; 1987 February 3- Crested Owls roost in the mid-canopy of the 7; Winnipeg, MB. Gen. T ech. Rep. RM-142. forest and Vermiculated Scr eech-owls in dense Fort Collins, CO: U.S. Department of Agri- shrubs. cultur e, Forest Service, Rocky Mountain Forest and Range Experiment Station: 135- At La Selva forest preserve, the Crested Owl, 143. Vermiculated Scr eech-owl, and Mottled Owl responded mor e to intra- and interspecific Carpenter, W. Thomas. 1987. Ef fect of envir on- broadcasting vocalizations; Black-and-white mental variables on r esponses of Eastern Owls responded less, and Spectacled Owls not Screech-owl to playback. In: Ner o, R.W.; at all. Dif ferences in r esponse levels to other Clark, R.J.; Knapton, R.J.; Hamr e, R.H., owl species depended on ecological variables eds. Biology and conservation of norther n such as habitat selection, population density, forest owls: symposium pr oceedings; 1987 and resource used. Our data suggest that the February 3-7; W innipeg, MB. Gen. T ech. development of r elationships within the tr opical Rep. RM-142. Fort Collins, CO: U.S. De- owl community at La Selva may have been partment of Agricultur e, Forest Service, mediated in part by intra- and interspecific Rocky Mountain For est and Range Experi- calling behavior. The interspecific r elationships ment Station: 277-280. could be a mechanism of habitat and r esource selection, and knowledge or these interactions Enríquez, Paula L. 1995. Abundancia r elativa, would be useful in developing management uso de hábitat y conocimiento popular de plans or conservation pr ograms (Mikkola los Strigifor mes en un bosque húmedo 1983). Habitat transfor mation in the sur - tropical en Costa Rica. Her edia, Costa Rica: rounding ar ea at La Selva has shown that owl Universidad Nacional. 81 p. M.S. disserta- abundances have varied thr ough time tion. (Enríquez 1995). In the Neotr opics, pr otected areas such as reserves and national parks Forsman, Eric D. 1983. Methods and materials function as r efuges for many species of wildlife for locating and studying Spotted Owls. that depend on for ested habitats (e.g., Cr ested Gen. T ech. Rep. PNW -162. Portland, OR: Owls). Finally, further infor mation on owl U.S. Department of Agricultur e, Forest behavior is r equir ed in order to better under - Service, Pacific Northwest Resear ch Station. stand the ecology of owl communities and 8 p. factors affecting owls in pr otected natural areas. Indeed, in tr opical areas habitat is the Ganey, Joseph. 1990. Calling behavior of most important factor to pr otect and to safe- Spotted Owls in Norther n Arizona. Condor . guard an owl community (Mikkola 1983). 92: 485-490.

ACKNOWLEDGMENTS Gerhardt, Richar d. 1991. Response of the Mottled Owl ( Ciccaba virgata) to br oadcast We thank Bruce Y oung (Or ganization for T ropi- of conspecific call. Jour nal of Field Or ni- cal Studies), Geor ge Rabb (Chicago Zoological thology. 62(2): 239-244. Society), the Costa Rican W ildlife Foundation, and the U.S. Fish and W ildlife Service (USFWS) Gerhardt, Richar d; Gerhardt, D.M.; Flatten, for financial support of this study. W e also C.J.; Bonilla, G.N. 1994a. The food habits thank Rosibel Barrantes for her field assis- of sympatric Ciccaba owls in Norther n tance. We thank Denver Holt, Richar d Guatemala. Jour nal of Field Or nithology. Gerhardt, and David Johnson for useful com- 65(2): 258-264. ments in this manuscript. Gerhardt, Richar d; Bonilla, G.N.; Ger hardt, D. LITERA TURE CITED M.; Flatten, C.J. 1994b. Br eeding biology and home range of two Ciccaba owls. Bosakowski, Thomas; Spelser, R.; Benzinger, J. Wilson Bulletin. 106(4): 629-639. 1987. Distribution, density, and habitat relationships of the Barr ed Owl in North Hekstra, G.P. 1973. Scops and scr eech owls. New Jersey. In: Nero, R.W.; Clark, R.J.; In: Owls of the world. London, England: Knapton, R.J.; Hamr e, R.H., eds. Biology Peter Lowe, Eur obook: 94-115. 531 2nd Owl Symposium Ibañez, C.; Ramo, C.; Busto, B. 1992. Notes on Mosher, James, A.; Fuller , M.R.; Kopeny, M. food habits of the Black-and-white Owl. 1990. Surveying woodland raptors by Condor. 94: 529-531. broadcast of conspecific vocalizations. Jour nal of Field Or nithology. 61(4): 453- Johnsgard, Paul. 1988. North American owls. 561. Washington, DC: Smithsonian Institution Press. 295 p. Nicholls, Thomas H.; W arner, D.W. 1972. Barred Owl habitat use as determined by Kochert, M.N. 1986. Raptors. In: Inventory and radiotelemetry. Jour nal of Wildlife Manage- monitoring of wildlife habitat. U.S. Depart- ment. 36: 213-225. ment of the Interior: 313-349. Pianka, Eric. 1988. Evolutionary ecology. New Korpimäki, Erkki. 1987. Composition of the owl York, NY: Harper & Row. Publishers. 468 p. communities in four ar eas in western Finland: importance of habitats and inter - Smith, Dwight; Devine, A.; W alsh, D. 1987. specific competition. In: Pr oceedings of the Censusing Scr eech Owls in souther n 5th Nor dic or nithological congr ess; 1985; Connecticut. In: Ner o, R.W.; Clark, R.J.; Acta Regiae Societatis Scientiarum et Knapton, R.J.; Hamr e, R.H., eds. Biology Litterarum Gothobur gensis Zoologica. 14: and conservation of norther n forest owls: 118-123. symposium pr oceedings; 1987 February 3- 7; Winnipeg, MB. Gen. T ech. Rep. RM-142. McGarigal, K.; Fraser , J.D. 1985. Barr ed Owl Fort Collins, CO: U.S. Department of Agri- responses to r ecorded vocalizations. Con- cultur e, Forest Service, Rocky Mountain dor. 87: 552-553. Forest and Range Experiment Station: 255- 267. Mikkola, Heimo. 1983. Owls of Eur ope. Calton, England: T. & A.D. Poyser. 397 p. Springer, M.A. 1978. Foot surveys versus owl calling surveys: a comparative study of two Mikkola, Heimo. 1992. W ood owls. In: Owls of Great Horned Owl censusing techniques. the world. Netherlands, Holland: Peter Inland Bir d Banding News. 50: 83-93. Lowe, Eurobook: 108-140. Stiles, Gary; Skutch, A. 1989. A guide to the of Costa Rica. Ithaca, NY : Cor nell University Pr ess. 511 p.

Will, Thomas C. 1986. The behavior ecology of species replacement: Blue-winged and Golden-winged warblers in Michigan. Ann Arbor, MI: University of Michigan. 126 p. Ph.D. dissertation.

Forest habitat at La Selva Biological Station, Costa Rica. Jose’ Luis Rangel Salazar

532