|
Article Excerpt
The Cloud-forest Screech Owl (Megascops marshalli) is one of the most poorly known owls in South America and worldwide. It is currently listed as Near Threatened under IUCN Red List criteria (BirdLife International 2008) and considered to be endemic to Peru. It occurs in mossy cloud forest, where it is apparently confined to a narrow altitudinal range of ~1,900-2,250 m (Weske and Terborgh 1981, Schulenberg et al. 1984, Fjeldsa and Krabbe 1990). The species was described quite recently (Weske and Terborgh 1981) based on eight specimens from Cordillera Vilcabamba in central Peru (12[degrees] 38' S, 73[degrees] 36' W). An additional bird was collected in 1982 in Cordillera Yanachaga, Peru (10[degrees] 37' S, 75[degrees] 20'W; Schulenberg et al. 1984), ~300 km northwest of the type locality (Fig. 1). It was considered to be common on Cordillera Vilcabamba, but the Cloud-forest Screech Owl is so elusive that neither of the above studies obtained a visually confirmed tape recording of the species' vocalizations.
SKH tape recorded an unidentified Megascops screech owl at 1,600 m (17[degrees] 10' S, 65[degrees] 35' W; Fig. 1) on 16 July 1996 during an altitudinal gradient study (Herzog et al. 2005) on Serranfa de Callejas above the village of El Palmar in the northwest corner of Bolivia's Parque Nacional Carrasco, Departamento Cochabamba. This species was tape recorded again at the same site ~2 years later by AM. Both recordings were included in Mayer (2000), but the owl's identity remained unknown.
[FIGURE 1 OMITTED]
Similar Megascops vocalizations were heard throughout August 2001 and tape recorded between 2,100 and 2,580 m on the east slope of Cordillera Cocapata in the Rfo Pampa Grande Valley (16[degrees] 41' S, 66[degrees] 29' W), Departamento Cochabamba, Bolivia (MacLeod et al. 2005; Fig. 1). We elicited vocal response of two individuals, a presumed pair, on 31 August 2001 using playback of the species' vocalizations recorded in this area. A male was lured into a mist net at 2,550 m by SKH and SRE shortly afterwards and prepared the following morning as a museum specimen by SKH. The bird was preliminarily identified in the field as a Cloud-forest Screech Owl based on the species' plumage description in Fjeldsfi and Krabbe (1990). A later comparison of the specimen with the description in Weske and Terborgh (1981) led to the same conclusion (MacLeod et al. 2005). However, given the often subtle differences in plumage patterns and coloration between species in the genus Megascops and the lack of vouchered or visually confirmed sound recordings of the species from the Peruvian type locality, this identification was tentative and required confirmation through a direct comparison of specimens.
We report on the specimen comparison, which confirmed the identification of the Bolivian specimen as a Cloud-forest Screech Owl, and present data on the species' vocalizations, distribution, natural history, and ecological relationships with sympatric Megascops screech owls in the Bolivian Yungas. We also reassess the conservation status of Cloud-Forest Screech Owl.
METHODS
Specimens were examined by SKH at the Louisiana State University Museum of Natural Science (LSUMZ). The plumage of the Bolivian specimen housed at the Coleccion Boliviana de Fauna (CBF 03760, male) was compared visually to two paratypes from Cordillera Vilcabamba sent on loan from the American Museum of Natural History (AMNH 824162, 824165; female and male, respectively) and the single specimen from Cordillera Yanachaga (LSUMZ 105658, male); digital photographs were taken for documentation. Measurements of wing chord were taken to the nearest 0.5 mm using a wing ruler; measurements of tail, bill (from anterior edge of nares), and tarsus length were taken to the nearest 0.1 mm using digital calipers.
In addition, SKH examined digital photographs of three specimens (1 male, 2 females) collected and identified as Cloud-forest Screech Owls by Carlos Mendoza (pers. comm.) and Jean Mattos on 3-5 November 2002 at Campamento Segakiato (a temporary gas pipeline construction camp; 12[degrees] 43' S, 73[degrees] 19' W; 1,850 m) in Provincia La Convencion, Departamento Cusco, Peru. Those specimens are housed at the Museo de Historia Natural in Lima (MUSM 25040-25042). SKH also compared several songs of an unidentified screech owl tape recorded by Hennessey (2004) on 26 June 2002 to our voucher recordings from Departamento Cochabamba; the owl was recorded in wet montane forest at Inciensal Sauce (14[degrees] 25' S, 68[degrees] 42' W; 2,300 m; Fig. 1) in Parque Nacional Madidi, Departamento La Paz, Bolivia.
Sound recordings of Bolivian Cloud-forest Screech Owls were obtained by SKH, SRE, AM, and A. B. Hennessey with Sony TCM 5000 EV cassette recorders and Sennheiser directional microphones. Recordings were digitized with an Edirol UA-1EX USB audio interface, and spectograms were examined and measured using Raven 1.2.1 (Charif et al. 2004) by SKH. Five recordings (2 natural, 3 after playback) from three localities (Appendix) were analyzed, representing at least four individuals and one vocalization type, the longsong. Assessment of relationships with other brown-eyed Andean Megascops was made by SKH by analyzing longsong recordings of Cinnamon Screech Owl (M. petersoni) from Ecuador and Peru, Rufescent Screech Owl (M. ingens) from Bolivia, Peru, and Ecuador, and of presumed Cloud-forest Screech Owl from Cordillera Yanachaga, Peru (Appendix). The identification of owls sound recorded on Cordillera Yanachaga in close proximity to the collecting locality of Schulenberg et al. (1984) as Cloud-forest Screech Owl was confirmed visually and independently by Jonas Nilsson (pers. comm.) on 6 June 2002 and TV on 8 March 2004, but no voucher specimen was collected.
Four variables were examined for each longsong: (1) duration (sec), measured from the beginning of the first to the end of the last note; (2) max frequency (Hz), i.e., the frequency at which maximum power (dB) occurred (Charif et al. 2004); (3) number of notes per second, i.e., the total number of notes divided by the duration of a song; and (4) change in pace, i.e., the duration (sec) of the last 10 notes and their respective internote intervals subtracted from the duration of the first 10 notes and their respective intemote intervals. We assessed the extent of similarity between longsongs of different species (and between regions within species) using principal component (PC) analysis and plotted factor scores of the second (PC II) against those of the first (PC I) principal component. PC I had an eigenvalue of 2.07. The eigenvalue of PC II was < 1.0 (0.87), and it was not used in further statistical analysis. PC I obtained moderate to...
|
